Investigating discharge communication for chronic disease patients in three hospitals in India.

OBJECTIVES: Poor discharge communication is associated with negative health outcomes in high-income countries. However, quality of discharge communication has received little attention in India and many other low and middle-income countries. PRIMARY
OBJECTIVE: To investigate verbal and documented discharge communication for chronic non-communicable disease (NCD) patients. SECONDARY
OBJECTIVE: To explore the relationship between quality of discharge communication and health outcomes.
DESIGN: Prospective study.
SETTING: Three public hospitals in Himachal Pradesh and Kerala states, India. PARTICIPANTS: 546 chronic NCD (chronic respiratory disease, cardiovascular disease or diabetes) patients. Piloted questionnaires were completed at admission, discharge and five and eighteen-week follow-up covering health status, discharge communication practices and health-seeking behaviour. Logistic regression was used to explore the relationship between quality of discharge communication and health outcomes. PRIMARY: Patient recall and experiences of verbal and documented discharge communication. SECONDARY: Death, hospital readmission and self-reported deterioration of NCD/s.
RESULTS: All patients received discharge notes, predominantly on sheets of paper with basic pre-printed headings (71%) or no structure (19%); 31% of notes contained all the following information required for facilitating continuity of care: diagnosis, medication information, lifestyle advice, and follow-up instructions. Patient reports indicated notable variations in verbal information provided during discharge consultations; 50% received ongoing treatment/management information and 23% received lifestyle advice. Within 18 weeks of follow-up, 25 (5%) patients had died, 69 (13%) had been readmitted and 62 (11%) reported that their chronic NCD/s had deteriorated. Significant associations were found between low-quality documented discharge communication and death (AOR = 3.00; 95% CI 1.27,7.06) and low-quality verbal discharge communication and self-reported deterioration of chronic NCD/s (AOR = 0.46; 95% CI 0.25,0.83) within 18-weeks of follow-up.
CONCLUSIONS: Sub-optimal discharge practices may be compromising continuity and safety of chronic NCD patient care. Structured protocols, documents and training are required to improve discharge communication, healthcare integration and NCD management.

Introduction

Continuity of care can most simply be defined as: “the seamless provision of healthcare between settings and over time”.[1] It is crucial for managing patients with chronic non-communicable disease (NCDs), who often require regular check-ups and care episodes across a variety of healthcare settings. Continuity of care is built around effective communication between healthcare professionals (HCPs) and between HCPs and patients. This is imperative during transitions of care, where poor handover communication can have far-reaching consequences such as delays in care, incorrect treatment and readmission.[2, 3]

Hospital discharge is one point of care transition that has proven to be particularly critical for continuity and safety of patient care. Evidence from high-income countries (HICs) has indicated that one in five patients experience an adverse event following discharge and that one-third of these events are preventable.[4] Regarding the impact of discharge communication for patient safety, there is an established link between the deficient exchange of documented information between hospital and primary care HCPs and adverse patient outcomes.[4-7] Ineffective communication between HCPs and patients/carers during discharge consultations, particularly regarding condition and/or ongoing treatment needs, has also been identified as an issue that can lead to patient misunderstanding and adverse events such as medication errors and unplanned readmissions.[4, 8–10] With regard to interventions, the literature has continually advocated the timely and accurate exchange of patient-specific information to improve coordination and safety during care transitions.[11, 12] Effective discharge education and documented summaries are particularly vital tools that have been shown to reduce a number of post-discharge complications and unplanned readmissions.[7, 13–18] However, in practice the delivery of discharge instructions often remains rushed and essential details for facilitating continuity of care such as diagnosis, medication, lifestyle and follow-up information are not always exchanged.[13, 19–24]

Despite the growing body of literature on quality of discharge communication and its impact on HCP and patient-related outcomes in HICs, similar evidence from low and middle-income countries (LMICs) remains relatively scarce.[25] Some single-site observational studies have evaluated discharge practices and found issues regarding deficient documentation, guidelines, standardised procedures and patient education.[26-30] In addition, a recent (2019) study from South Africa found inadequate discharge planning (a process which involves healthcare information transfer between HCPs and between HCPs and patients/carers to ensure coordination and continuity of care) to be a significant contributor to potentially avoidable causes of readmissions.[31] Across India, a handful of studies have found inconsistencies in the provision of discharge information via HCP and patient reports and discharge ticket evaluations.[32-34] In addition, a study in an Indian hospital emergency department reported improvements in recorded discharge information following the implementation of pre-formatted discharge summaries.[35]

The importance of investigating factors affecting continuity of care in LMICs and India, in particular, is increasing due to the rising prevalence of chronic NCDs, which require sustained care across settings.[36] Given the resource constraints across numerous LMIC settings, the need to elucidate context-relevant strategies to improve patient self-management and avoid unnecessary healthcare utilisation is also vital. In the study areas of India (i.e. Himachal Pradesh and Kerala states), a lack of standardised information systems within and between levels of healthcare has resulted in patient-held medical information serving as the predominant vehicle for handover communication between HCPs and between HCPs and patients.[37, 38] As well as verbal advice/instructions, patients are often provided with medical documents (including discharge notes) during healthcare visits that can be transported between HCPs whilst seeking care from a variety of public and private providers. Therefore, the quality of both verbal and documented information exchanged between HCPs and patients is critical in facilitating adequate comprehension, coordination and continuity of care.

We conducted a prospective cohort study with the primary objective of investigating verbal and documented discharge communication for chronic NCD patients in three hospitals in Himachal Pradesh and Kerala states, India. In addition, given the increasing evidence on the significance of information exchange for patient outcomes, a secondary objective was to explore the relationship between quality of discharge communication and adverse health outcomes.

Methods

Study setting

Overview

The study was conducted from December 2014 to November 2015 in Himachal Pradesh and Kerala states. Patients were recruited and initial data was collected from three hospitals: one rural secondary-care hospital (150 beds) in the district of Solan, Himachal Pradesh, and one peri-urban secondary-care hospital (150 beds) and one urban tertiary-care hospital (783 beds) in the district of Ernakulam, Kerala. These settings were selected to capture a range of contrasting hospital types and environments. We selected public rather than private healthcare facilities for this study as this is where targeted improvements to health systems could be implemented more systematically (i.e. via state health departments). In addition, public facilities are where a large proportion of socio-economically vulnerable populations access healthcare in India.

Public healthcare in India

At the national level, public healthcare in India is directed by the Ministry of Health and Family Welfare. At the state level, public healthcare is managed by the State Department of Health and Family Welfare, which has considerable autonomy in deciding upon, designing and delivering health programs. Fig 1 contains a summary of the structure of public healthcare in India, based on Indian Public Health Standard Norms.[39]

Fig 1

Structure of the Indian public healthcare system according to Indian Public Health Standard Norms.

Population and public healthcare in Solan, Himachal Pradesh and Ernakulam, Kerala

Himachal Pradesh is a predominantly rural and mountainous state in northern India with a population of 6.86 million people.[40] Solan district has a population of approximately 580,000 people, with 82.4% living in rural areas. There average literacy rate across the district is 83.7%, which is higher than the national average (74%) but rates remain lower for women compared to men (77.0% vs. 89.6%, respectively).[41] A recent study regarding the availability of health services across Himachal Pradesh found that there are 5 hospitals, 6 Community Health Centres (CHCs), 33 Primary Health Centres (PHCs) and 179 Sub Health Centres (SHCs) in Solan District. With regard to primary care infrastructure, it was calculated that there are approximately 0.83 CHCs per 80,000 persons, 1.14 PHCs per 20,000 persons and 0.93 SHCs per 300 persons.[42]

Kerala is a state in the south-west of India with a population of 34.8 million people and a greater than national average urban-based population of 47.7%.[43] Ernakulam district has a population of approximately 3.2 million people, with 68.1% living in urban areas. The average literacy rate across the district is 95.89%, which is notably higher than the national average with similar rates between women and men (94.5% vs. 97.4% respectively).[44] Regarding the availability of health services, the Kerala Department of Health Services (DHS) reports there are 15 hospitals, 23 CHCs, 75 PHCs and 410 SHCs in Ernakulam district.[45] Based on census and DHS data, it can be calculated that there are approximately 0.58 CHCs per 80,000 persons, 0.47 PHCs per 20,000 persons and 0.04 SHCs per 300 persons in Ernakulam.

Ethics approval

This study was reviewed and approved by the Centre for Chronic Disease Control Independent Ethics Committee, India, and the Amrita Institute of Medical Sciences Institutional Ethics Committee, India. Data archives are stored at the University of Birmingham, in accordance with the University’s code of practice.

Patient recruitment

Patients were recruited consecutively by trained social work graduate researchers (n = 6) six days per week (spread across all days of the week over the study period) between the hours of 8 am and 6 pm, as this is the window within which patients were typically discharged from study hospitals. Patients were included within the first 24 hours of their admission if they met the following inclusion criteria: adults (18yrs+) with one of the following tracer chronic NCDs; cardiovascular disease, chronic respiratory disease, diabetes mellitus or hypertension. Researchers approached ward nurses to identify eligible chronic NCD patients; patients were excluded they lived outside the study areas (where planned follow-up visits would be difficult to achieve), or if ward nurses judged them to be too unwell to participate due to physical and/or cognitive impairment. Researchers then approached eligible patients (and their carers), provided them with study information sheets and verbally informed about the purpose of the research. Written consent was obtained from literate patients. For illiterate patients, oral consent was obtained along with a thumbprint and signature from a literate witness (i.e. carer) in line with World Health Organisation ethical guidelines.[46]

Due to the lack of prior work completed in the field of study, the sample size for this study was calculated using a formula to determine the minimum sample size needed to estimate a population mean with confidence limits of 5% for a variable (such as the proportion of patients receiving complete healthcare information) with a prevalence of 50%. The formula used was: Necessary Sample Size = (Z-score)2 * StdDev*(1-StdDev) / (margin of error)2.[47] Based on our parameters this worked out as 384.16. Therefore, we aimed to collect survey data from a minimum of 385 patients.

Data collection

Questionnaire

The questionnaire used in this study was developed from one used in previous handover research conducted in Nigeria for a masters in public health thesis (see “S1 Appendix” for a copy of the questionnaire). It was adapted using relevant handover communication literature and input from expert researchers in the UK and India. Prior to the commencement of data collection, a small pilot study was conducted in Kerala to test the questionnaire. This was conducted iteratively over three rounds, with field researchers visiting the study hospitals and interviewing two patients/carers at each location. In addition to asking questions, researchers also asked for feedback on the clarity and contextual appropriateness of the materials. Once this was completed, all researchers met with a supervisor to discuss the aim of each question, patient/carer responses and subsequently make minor amendments to the wording and structure of the questionnaire to improve clarity. After three rounds of piloting, based on both patient/carer and researcher feedback, the questionnaire was considered to be suitable for use; piloted cases were not included in the main study.

Questionnaires were completed at four time points: 1. during hospital admission; 2. immediately after hospital discharge; 3. five weeks after discharge and 4. eighteen weeks after discharge. All questionnaires were administered in person by trained social work graduate researchers (n = 6) in the form of an interview; each question was read aloud and researchers ticked the appropriate box/es for responses that corresponded to pre-defined answers and/or wrote free-text notes for responses that did not correspond to pre-defined answers. Whilst patients predominantly provided responses to questions, their accompanying carer/s (e.g. friends, family members etc.) were able to provide support for clarifying answers when required. The admission and discharge interviews were completed in the hospital, whereas the five-week and eighteen-week follow-up interviews were completed in the community.

During admission, data was collected on patient demographics, health status, previous HCP visits, referral/s (if any) and experiences of healthcare information transfer. Following hospital discharge consultations, information was collected about verbal recall of healthcare information given to patients, attitudes towards such information and subsequent plans for HCP follow-up. Researchers also assessed patient/carer understanding of their health condition and post-discharge care requirements by asking them to explain this information and then checking it against their discharge notes, or if these did not contain enough information/were not immediately available, a ward nurse who was aware of their condition and care requirements. Researchers also evaluated the contents of all notes, prescriptions and/or other pieces of paper that the patient had brought with them to hospital and that had been given to them by a hospital HCP during admission and/or their discharge consultation. These content evaluations were completed using structured checklists within the questionnaires; space was provided for free-text entry to cover any information not included in the checklists. At five and eighteen-week follow-up visits data was collected about post-discharge adverse health outcomes, health status and health-seeking behaviour.

Analysis

Descriptive statistics were used to outline patient demographics, attitudes and experiences of discharge communication and post-discharge health outcomes. Descriptive statistics were also used to calculate the number of patients who received all items of “key” documented and verbal information during discharge communication. For documented information, key items included: Diagnosis, medication information, lifestyle advice and follow-up instructions. For verbal information, key items included: Ongoing treatment/management information, medication information, lifestyle advice and follow-up instructions. These items were selected based on common themes across the literature regarding critical information needed to improve discharge communication and care transitions for patients with chronic conditions.[48-51]

Due to a relatively small number of outcome events in the study sample, the Firth method of multiple logistic regression was used to explore the relationship between quality of discharge communication and health outcomes at five and eighteen-weeks follow-up.[52, 53] The principle outcome variables were death (all-cause mortality), hospital readmission and self-reported deterioration of NCD/s (that the patient was hospitalised for). For explanatory variables, the completeness of key documented and verbal information was utilised to measure the quality of discharge communication. The explanatory variables of interest in this study were receiving low-quality documented discharge communication (i.e. notes containing 0–2 items of key documented information—compared to notes containing 3–4 items) and receiving low-quality verbal discharge communication (i.e. recall of receiving 0–2 items of key verbal information—compared to recall of receiving 3–4 items) during discharge consultations.

Multiple logistic regression models were employed in order to adjust for the following potential confounders: sex, age, education level, employment status, time taken to reach the hospital, number of chronic NCDs and hospital site. A count of chronic NCDs was used due to a lack of available data regarding primary diagnosis or severity of comorbidities, which would have enabled the creation of a comorbidity index. Such an approach has been validated and shown to add predictive value for survival analyses (vs. age alone).[54]

Results

Demographics and medical conditions

A total of 546 inpatients completed questionnaires; 305 men and 241 women. The majority of participants were aged 60 years or older (59%) and were literate with a complete primary school-level education or more (67%). Of the four chronic NCDs captured by this study, the most frequently reported was chronic respiratory disease (45%) (Table 1). See “S2 Appendix” for participant demographic information and health outcomes by study site.

Table 1

Participant demographic and adverse health outcomes information.

Characteristic	Total (n = 546)
	Frequency (%)
Sex
Male	305 (55.9)
Female	241 (44.1)
Age group (Years)
18–49	98 (17.9)
50–69	296 (54.2)
≥70	152 (27.8)
Level of education
Illiterate	91 (16.7)
Literate with partial or completed primary school education	258 (47.3)
Complete secondary school education	132 (24.2)
Complete higher school/vocational studies	52 (9.5)
University graduate or above	13 (2.4)
Employment status
Employed	164 (30.0)
Unemployed	369 (67.6)
Retired	11 (2.0)
No data*	2 (0.4)
Time taken to reach hospital
<1 hour	311 (57.0)
1–4 hours	230 (42.1)
>4 hours	4 (0.7)
No data*	1 (0.2)
Chronic NCDs†
Diabetes	157 (28.8)
Cardiovascular Disease	218 (39.9)
Chronic Respiratory Disease	247 (45.2)
Hypertension	171 (31.1)
Number of chronic NCDs (per patient)
1	365 (66.9)
2	128 (23.4)
3	40 (7.3)
4	13 (2.4)
Adverse health outcomes at 5-week follow-up
Death (all-cause mortality)	19 (3.5)
Hospital Readmission	33 (6.0)
Self-reported deterioration of NCD/s	39 (7.1)
No data* (loss to follow-up)§	13 (2.4)
Adverse health outcomes at 18-week follow-up
Death (i.e. all-cause mortality)	25 (4.6)
Hospital Readmission	69 (12.6)
Self-reported deterioration of NCD/s	62 (11.4)
No data* (loss to follow-up)§	14 (2.6)

* No data = missing responses

†Please note that participants could select more than one answer for this question

§Patients lost to follow-up were those who could not be contacted or found during community visits in the follow-up period

Hospital discharge

Patient recall of verbal discharge communication

Most patients (89%) reported having their health condition explained to them during admission. Post-discharge care advice appeared to vary notably, as just over half (50%) of all patients recalled being given information regarding ongoing treatment/management and 23% of patients recalled receiving lifestyle advice. With regard to follow-up instructions, the majority of patients (85%) recalled being told to return for an outpatient check-up. Overall, just 15 (3%) patients recalled receiving all key verbal discharge information (i.e. ongoing treatment/management information, medication information, lifestyle advice and follow-up instructions). With regard to patient understanding of post-discharge care information, a quarter of patients/carers (25%) were judged by researchers to have a good understanding of almost all important details (Table 2).

Table 2

Patient recall of verbal discharge communication and follow-up plans.

Hospital discharge	No. (n = 546)	%
Health condition explained to patient during admission
Yes	485	88.8
No	38	7.0
Patient unsure	16	2.9
No data*	7	1.3
Post-discharge care information provided by hospital doctors†
Instructions to go for further test/s	19	3.5
Details regarding ongoing management	274	50.2
Details of prescribed course of medication to be taken (and reviewed when completed)	297	54.4
Lifestyle advice (i.e. regarding exercise, diet, tobacco and/or alcohol)	123	22.5
Instructions to take rest	1	0.2
Instructions to visit a physiotherapist	14	2.6
Referral to another HCP	3	0.5
Patient unsure of what advice was given (if any)	12	2.2
No advice given	1	0.2
No data*	10	1.8
Follow-up instructions provided by hospital doctors†
Visit/s to the outpatient department of this (same) hospital	430	78.8
Visit/s to the outpatient department of another hospital/specific doctor	33	6.0
Patient unsure of what advice was given (if any)	75	13.7
No data*	16	2.9
Patient recalled receiving all key verbal discharge information§	15	2.7
Patient/carer understanding of health condition and post-discharge care requirements
Patient/carer had a good understanding of almost all important information	135	24.7
Patient/carer had a broadly correct understanding of important information	186	34.1
Patient/carer had only a basic understanding of some important information (e.g. diagnosis/medicine)	85	15.6
Patient/carer had very little/no understanding of any important information	20	3.7
No data**	120	22.0
Patient plans for follow-up HCP visit/s†
Return to same hospital outpatient clinic	448	82.1
Another government hospital	12	2.2
Government primary care centre	20	3.7
Private hospital/nursing home	6	1.1
Local private doctor/nurse	6	1.1
Physiotherapist	1	0.2
Traditional healer	1	0.2
Patient plans to only return to a HCP when they are sick again	68	12.5
No data*	8	1.5
How patients plan to explain to next HCP what was done for them during admission†
Patient asked the doctor to explain to their carer/family member	247	45.2
Patient asked the doctor to explain to them so they can tell the HCP when they see them	82	15.0
Doctor gave the patient/carer a note or discharge summary to take back to their local HCP	116	21.3
Patient plans to return to this (same) hospital where their medical records are stored	100	18.3
Patient is unsure what they will do because they cannot remember what the doctor said	2	0.4
No data*	23	4.2

* No data = missing responses

† Participants could select more than one answer for this question

§ Ongoing management information, medication information, lifestyle advice and follow-up information

**No data = missing responses; please note the larger number of missing responses to this question was due to a lack of available documented discharge information and/or ward nurses, which were required at the time of questioning for the researcher to verify patient/carer understanding of their condition and post-discharge care requirements

Patient follow-up plans

When asked about follow-up plans, the majority of patients (82%) stated that they planned to return to the outpatient clinic of the same hospital they were being discharged from. A notable proportion of patients also (13%) stated that they would only return to a HCP when they become unwell again (Table 2). When patients were asked about how they would explain to the next HCP what was done for them during admission, the most common response (45%) was that they had asked the doctor to explain key information to their carer/family member (Table 2).

Documented discharge communication

All patients were provided with a document containing handwritten notes during their discharge consultation. Overall, most patients (94%) felt it was important to receive discharge notes, the most common reason given for this was because it helps patients to understand and explain their condition (64%). The type of discharge documents provided to patients varied; the majority received either a sheet of paper with basic pre-printed headings (71%—see “S3 Appendix” for a photographed example) or an unstructured (i.e. otherwise blank) sheet of paper (19%). The contents of discharge notes received by patients varied greatly between individuals, with only 31% containing all four items of key healthcare information required for effectively facilitating transitions of care (i.e. diagnosis, medication information, lifestyle advice, and follow-up instructions) (Table 3).

Table 3

Documented discharge communication.

Hospital discharge	No. (n = 546)	%
Documented information given to patients at discharge
Patient received discharge document/s (seen by a researcher)	546	100
Patient attitudes regarding importance of receiving discharge document/s
It is important to receive discharge document/s	513	94.0
It is not important to receive discharge document/s	24	4.4
Unsure whether it is important or not to receive discharge document/s	6	1.1
No data*	3	0.5
Reasons given for why patients feel it is important†	No. (n = 513)	%
It helps to understand and explain my condition/s	330	64.3
It helps me to get attended to faster at my next HCP visit	85	16.6
I feel it’s more professional	19	3.7
I have to submit this for claiming insurance	89	17.3
It will help in an emergency	2	0.4
It is a helpful medical identification certificate	1	0.2
Reasons given for why patients feel it is not important†	No. (n = 24)	%
The notes get lost	5	20.8
Everyone receives the same standard of care regardless	19	79.2
Types of documents given to patients at discharge
Discharge booklet	42	7.7
Structured discharge document (i.e. form/card with basic pre-printed headings)	386	70.7
Unstructured discharge document (i.e. note/letter on otherwise blank sheet of paper)	104	19.0
Prescription card (containing medication information only)	5	0.9
Referral letter	1	0.2
No data*	8	1.5
Contents of discharge documents	No. (n = 545)§	%
Illegible notes	29	5.3
Name of doctor/contact at the hospital	379	69.5
Date	517	94.9
Name, age and sex of patient	523	96.0
Diagnosis	536	98.4
Medication information	477	87.5
Follow-up instructions	299	54.9
Lifestyle advice (e.g. exercise, diet, tobacco, alcohol etc.)	268	49.2
Past medical history for current condition	331	60.7
Past medical history for other conditions	99	18.2
Patient’s signs, symptoms and problems when admitted	506	92.8
Tests performed during admission (without results)	98	18.0
Tests performed during admission (with results)	429	78.7
Discharge document contained all key items of information**	168	30.8

* No data = missing responses

† Participants could select more than one answer for this question

§ One patient did not give permission for the contents of their discharge document/s to be analysed

** Diagnosis, medication information, lifestyle advice and follow-up instructions

Five-week follow-up

Five weeks post-discharge 13 (2%) patients had been lost to follow-up, 19 (4%) patients had died, 33 (6%) patients reported they had been re-admitted to hospital and 39 (7%) patients reported a deterioration in their health (related to the NCD/s they were hospitalised for) (Table 1). See “S4 Appendix” for flowchart summarizing participant inclusion and exclusion throughout the study.

Results of the adjusted analyses of the association between low-quality discharge communication and death, hospital readmission and self-reported deterioration of NCD/s within 5 weeks of discharge are presented in Table 4 (see “S5 Appendix” for unadjusted results). They showed significant (p<0.05) increased odds of death within five weeks of discharge for patients who received low-quality discharge notes (AOR = 4.43; 95% CI 1.46, 13.46). No other significant associations were found. Goodness-of-fit tests using the method of Heinze and Schemper were performed for each adjusted multivariate analysis and are reported in “S6 Appendix”. In this approach, coefficients are constrained to zero and left in the model in order to allow their contribution to the penalization.[55]

Table 4

Adjusted associations between receiving low-quality discharge communication and the likelihood of experiencing adverse health outcomes within five and eighteen weeks of discharge.

Death within 5 weeks of discharge	Adjusted odds ratios*			Death within 18 weeks of discharge	Adjusted odds ratios*
	OR	95% CI	p-value		OR	95% CI	p-value
No. of items of key documented discharge information†				No. of items of key documented discharge information†
0 to 2 items	4.37	1.46–13.11	0.009§	0 to 2 items	3.00	1.27–7.06	0.012§
No. of items of key verbal discharge information**				No. of items of key verbal discharge information**
0 to 2 items	3.18	0.15–67.18	0.458	0 to 2 items	1.60	0.37–6.89	0.525
Hospital readmission within 5 weeks of discharge				Hospital readmission within 18 weeks of discharge
No. of items of key documented discharge information†				No. of items of key documented discharge information†
0 to 2 items	0.60	0.25–1.43	0.252	0 to 2 items	0.75	0.42–1.34	0.324
No. of items of key verbal discharge information**				No. of items of key verbal discharge information**
0 to 2 items	0.73	0.30–1.75	0.479	0 to 2 items	0.82	0.45–1.50	0.510
Self-reported deterioration of NCD/s within 5 weeks of discharge				Self-reported deterioration of NCD/s within 18 weeks of discharge
No. of items of key documented discharge information†				No. of items of key documented discharge information†
0 to 2 items	1.41	0.70–2.86	0.338	0 to 2 items	1.59	0.89–2.85	0.115
No. of items of key verbal discharge information**				No. of items of key verbal discharge information**
0 to 2 items	0.65	0.32–1.34	0.241	0 to 2 items	0.46	0.25–0.83	0.010§

*Adjusted for the following independent variables: sex, age group (18-49/50-69/70yrs+), education level (up to primary school-level/secondary school-level/higher school-level or more), employment status (unemployed/employed/retired), usual time taken to reach hospital (<1 hour/1-4 hours/>4 hours), number of chronic NCDs (1/2/3/4) and hospital site (1/2/3).

† Odds ratios represent association with receipt of 0 to 2 items of key documented information on discharge notes

§ Statistically significant at p<0.05

** Odds ratios represent association with receipt of 0 to 2 items of key verbal information during discharge consultation

Eighteen-week follow-up

Eighteen weeks post-discharge 14 (2%) patients had been lost to follow-up, 25 (5%) patient had died, 69 (13%) patients reported that they had been readmitted to hospital and 62 (11%) patients reported a deterioration in their health (related to the NCD/s they were hospitalised for) (Table 1).

Results of the adjusted analyses of the association between low-quality discharge communication and death, hospital readmission and self-reported deterioration of NCD/s within eighteen weeks of discharge are presented in Table 4 (see “S6 Appendix” for unadjusted results). They showed significant (p<0.05) increased odds of death for patients who received low-quality discharge notes (AOR = 3.00; 95% CI 1.27, 7.09). With regard to verbal information, the results showed a significant (p<0.05) decreased odds of self-reported deterioration of NCD/s within eighteen weeks of discharge for patients who recalled receiving low-quality verbal discharge communication (AOR = 0.48; 95%CI 0.27–0.87). No other significant associations were found (see “S6 Appendix” for goodness-of-fit statistics for all adjusted analyses).

Discussion

Main findings

This study has described discharge communication for chronic NCD patients in three public hospitals in India. One of our main findings was that both verbal and documented communication between HCPs and patients was often limited. Regarding documented information, discharge notes were predominantly provided on forms within minimal structure and in some cases were hastily written on blank sheets of paper or prescription forms. Patient attitudes towards discharge documents were generally positive, with the majority feeling it was important to receive them to help with understanding/explaining their condition, claiming insurance and getting attended to faster when they see the next HCP. However, the contents of discharge notes varied notably, with only just under a third of patients received notes containing all items of key information (i.e. diagnosis, medication information, lifestyle advice and follow-up instructions).

In addition, whilst the majority of patients reported being told where to go for a follow-up HCP visit during discharge consultations, other verbal information appeared to vary notably between patients. In particular, only just over half of all patients recalled receiving information about necessary ongoing treatment/management for their NCD/s and under a quarter recalled receiving lifestyle advice. A markedly small minority of patients (3%) recalled receiving all four items of key verbal information during discharge consultations (i.e. ongoing treatment/management information, medication information, lifestyle advice and follow-up instructions). In addition, just under a fifth of patients were assessed by researchers as having either a basic or little/no understanding of important discharge information and 13% of patients only planned to return to a HCP when they were unwell again. These findings indicate that a significant proportion of patients left the hospital with sub-optimal levels of discharge information and/or comprehension regarding their ongoing healthcare needs, which may have compromised their ability to adequately manage their chronic NCD/s. Given the role that patient-held documents play in facilitating handover communication between HCPs in the study areas of India, the deficiencies in documented information provision may have also affected continuity of patient care.

Overall, the results reflect the limited similar research from India that has evidenced unstructured and deficient HCP-patient communication at the point of discharge.[32, 35] They are also consistent with other LMIC-based studies that, via patient reports and record evaluations, have evidenced a lack of in-depth information provision during discharge and/or poor levels of patient understanding regarding post-discharge care requirements.[27-30] The provision of deficient documented discharge information may be of particular concern for patient self-management, as global literature (predominantly from high-income countries) has indicated that individuals can struggle to absorb the verbal information provided by HCPs during healthcare consultations.[56] Further, whilst there is a dearth of empirical research on HCP attitudes towards discharge in India, our research on outpatients and qualitative data regarding inpatients from the same study areas in India indicates that a paucity of time, available HCPs, training and guidelines are likely to be notable contributors to the suboptimal communication evidenced in this study.[37] The results may also be somewhat explained by a historical lack of communication training in medical education and a tendency for paternalistic physician behaviour in India.[57, 58] A dominant HCP communication style is likely to result in unmet information needs from patients, due to them feeling intimidated and unable to ask questions. Similar communicative issues have been identified across healthcare settings in other parts of Asia, despite an increasing desire from patients for more involvement.[59] A global review of discharge communication literature suggests that both patients and HCPs prefer practices that are relevant, concise and personalised.[60] However, in practice HCPs across numerous healthcare settings report not having enough time to perform comprehensive discharge consultations and instead prioritising inpatient medical care.[24, 61] Overall, the international evidence would suggest that the issues found in this study pose a serious challenge for the Indian public health system, given the importance of information exchange in ensuring the continuity and safety of healthcare.[62]

Furthermore, to the best of the authors’ knowledge, this study is the first of its kind to investigate the association between quality of discharge communication and adverse health outcomes in India. In the adjusted analyses we found that chronic NCD patients who received low-quality discharge notes (i.e. containing 0–2 items of key information) were more likely to have died within five and eighteen weeks of follow-up compared to those who received higher-quality notes. Such findings reflect HIC research, which has repeatedly demonstrated a link between deficient discharge information transfer and an increased risk of adverse events.[4-7] We also found that patients who received low-quality verbal discharge communication were less likely to report an NCD-related deterioration in their health within eighteen weeks of follow-up compared to those who received higher-quality communication. This was an unexpected finding, which could be explained by the possibility that those provided with less information at discharge were patients with less severe health issues (due to HCP time pressures/case prioritisation etc.) and were subsequently less likely to report a deterioration later on. Overall, it is not possible to say whether low-quality discharge communication caused some, or all, of the health outcomes observed in this study. The small scale of the research alongside the complex nature of factors affecting health outcomes means that the results should be interpreted cautiously and will require further validation. Nonetheless, given the importance of handover communication for continuity of safety of care, the imperative to improve the recording and transfer of key healthcare information remains.

Strengths and limitations

A key strength of this study is the collection of data from a range of patients across multiple healthcare facilities in varying geographical settings, which has provided a representative sample of chronic NCD patients accessing public hospitals in two diverse states of India. This study has also provided key insight into the transfer of critical NCD patient information during public hospital discharge in India. Follow-up data can be challenging and resource-intensive to collect in LMIC settings for a number of reasons, including a lack of comprehensive/up-to-date directories to assist with locating addresses, difficulties in contacting individuals to confirm visit attendance (due to a lack of phone ownership/poor network coverage) and limited access to certain areas (due to challenging terrains and/or limited transportation infrastructure). Therefore, this study has also provided an invaluable opportunity to explore the experiences of NCD patients following hospital discharge in the community.

However, given the vastness of India and the complexity of healthcare systems across LMICs, generalisability to other settings may be done with caution. A limitation of this study is that the quality of discharge communication was predominantly assessed via patient recall, rather than direct observation. A lack of adequately recorded inclusion/exclusion rates for participation is also a limitation as this could not be reported. The involvement of six different researchers may have increased the likelihood of researcher bias and, as the same researchers completed all data collection, they were not blinded to the quality of discharge communication. However, each section of the questionnaire was immediately filed away after completion and not referred to again which, alongside the fact that each researcher collected large volumes of questionnaire data, reduced potential for further bias. Regarding limitations of the regression analyses, data regarding diagnostic accuracy was unavailable and, theoretically, associations may have arisen as a result of imprecision (alpha error), indirect causal links (e.g. the provision of information was a marker for other aspects of care) or a combination of these factors. The limited number of deaths means it is also unclear how well our models adjusted for confounding factors, so the findings must be interpreted with caution. In addition, our follow-up questionnaire did not capture further information on patients who had died, so it was not possible to know whether these patients had been readmitted. Therefore, we ran further sensitivity analyses based on the assumption that all patients who died had also been readmitted; whilst no significant associations were found, adjusted point estimates regarding associations with low-quality discharge notes all leaned in the direction of an increased likelihood of hospital readmission within five and eighteen weeks follow-up (see “S7 Appendix” for results and goodness-of-fit statistics).

Next steps

To effectively address the issues surrounding ineffective communication between HCPs and patients during discharge consultations, there is a pressing need for structured HCP training and handover communication guidelines to be implemented across public healthcare facilities–our outpatient handover research from the same study hospitals has indicated that these are currently missing.[63]

Extensive HIC-based research has shown that education, simulation-training, and communication tools are effective strategies for improving the quality of healthcare information transfer during transitions of care.[12] Communication tools can also be adapted to ensure they facilitate patient-centred care at the point of discharge, which takes into account patients’ needs, desires and values and involves patients and carers in care/decision-making processes [12, 24] In addition, several LMIC studies have evidenced that introducing patient discharge educational materials and structured disease-specific discharge planning can improve HCP to patient communication, as well as patient satisfaction, patient/carer healthcare management knowledge and post-discharge health outcomes.[64-68] Given the predominantly paper-based systems in use across the study settings, well-structured and standardised HCP checklists, documents and patient-held record booklets are also likely to advance the quantity and quality of essential information transferred between HCPs and between HCPs and patients and have proven successful in HIC and LMIC settings.[35, 69–71] Co-creation of such materials with HCPs, patients, carers and other key stakeholders should be considered in order to enhance acceptability, function and utilisation.

Looking to the future, the introduction of electronic health information systems holds great promise for improving information exchange and overall quality of healthcare.[72] A review of computer-enabled discharge communication interventions has evidenced their impact in improving timeliness and accuracy of information as well as patient and HCP satisfaction across Europe and North America.[60] In India, the government has recently announced long-term plans to digitize health records and this is currently being set up in seven states, including Himachal Pradesh and Kerala.[73] In addition, Kerala is the first Indian state to undergo comprehensive e-health systems reforms across all public healthcare facilities.[74] However, these developments are presently in their early stages and will take some time to become integrated enough to effectively facilitate handover communication; patient-held medical documents remain in use as the predominant vehicle for information transfer throughout public healthcare. In addition, e-health developments will not address integrative challenges between public and private HCPs using different information systems. This may continue to compromise continuity of care for many patients visiting multiple HCPs. Further, technological advances will not necessarily address improvements regarding the quality of information exchanged between HCPs and patients.

Conclusion

In conclusion, this study has found that the quality of discharge communication for chronic NCD patients visiting public healthcare facilities in Himachal Pradesh and Kerala states in India is currently suboptimal. As a consequence of this, many patients with ongoing health needs are leaving hospitals with insufficient levels of information and/or understanding to be able to facilitate continuity of care and adequately manage their NCD/s. The findings have also evidenced significant associations between low-quality discharge communication and health outcomes at five and eighteen weeks follow-up. Whilst these associations must be interpreted with caution, overall this study has highlighted a pressing need for the wide-scale implementation of structured protocols, documents and training to improve the exchange of key discharge information between HCPs and between HCPs and patients. With the rising burden of NCDs across India and other LMICs, the findings from this study are timely and crucial for effective health systems development. Regarding future research, additional in-depth investigation is required to elucidate the validity of relationships between discharge communication and health outcomes. It is also important that further robust LMIC studies are conducted to continue exploring the critical factors affecting the continuity and safety of NCD patient care and develop sustainable, cost-effective interventions.

Copy of previous handover research study questionnaire used as basis for current study questionnaire.

(PDF)

Click here for additional data file.

Baseline characteristics and adverse health outcomes by study site.

(PDF)

Click here for additional data file.

Exemplar picture of a structured discharge slip.

(PDF)

Click here for additional data file.

Summary flowchart of participant inclusion and exclusion throughout the study.

(PDF)

Click here for additional data file.

Results of all unadjusted regression analyses.

(PDF)

Click here for additional data file.

Summary of goodness-of-fit test results for all adjusted multivariate analyses.

(PDF)

Click here for additional data file.

Summary of results and goodness-of-fit test results for all sensitivity analyses.

(PDF)

Click here for additional data file.

11 Dec 2019

PONE-D-19-22862

INVESTIGATING CHRONIC DISEASE PATIENTS’ ATTITUDES AND EXPERIENCES OF DISCHARGE COMMUNICATION IN THREE HOSPITALS IN INDIA

PLOS ONE

Dear Sir,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. The paper deals with an interesting topic and we feel that it merits publication.  Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. Please adress the major concerns especially raised by the reviewer 2.

==============================

We would appreciate receiving your revised manuscript by 13/01/202.  When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Prof. Prabath  Nanayakkara, MD, PhD, FRCP

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include additional information regarding the PILOT questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, how many participants were included in this pilot testing and was the original questionnaire altered in any way following the trial?

3. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: I Don't Know

Reviewer #3: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: In general, I think this article is well-written and covers an important topic of quality of medical care. The authors investigated whether discharge information is appropriate provided and investigated associations between the handover quality and health outcomes.

Major comments:

I am not sure whether the discovered associations between the (experienced) quality of the handover and health outcomes (especially all cause mortality) are that relevant, but these are a big part of the manuscript.

As the authors also state in their discussion this relation has to be interpreted with caution as confounding bias (in my opinion) may play an important role. For instance, compliance to therapy has not been measured and I can only imagine the quality of handover being relevant in mortality if this is caused by an Adverse Event of the specific disease (instead of other causes).

I think it's up to the editor to decide if this needs a change in the manuscript.

Minor comments

- Regarding the methods oft his study I think it's a limitation not observing the verbal handover information. Now, the patients' recall is measured, which of course indicates how well patients understand the provided information; however, there might be some differences between what has been told and what is recalled by patients, which may have implications for quality improvement strategies. In addition, the teach-back method in my opinion could have been a better method instead of using multiple choice questions to measure recall.

- Concerning exlusion: are patients with cognitive impairment included as well (not stated they have been excluded)

- You may consider to describe the pilot study a bit more in detail: how many patients were involved? Or did only researchers optimize the questionnaire? In that case, I think the study could have been optimized with pilot testing/cognitive interviewing of patients

- Abbreviations do come up in tekst without previous being presented in (..). For example LMIC line 114, CHC/PHC/SHS line 171

- Can you add a reference to the questionnaire used as a basis?

Reviewer #2: SUMMARY

This study describes the discharge communication process in association with adverse health outcomes in India. It was a prospective study including non-communicable patients in three public hospitals. The study shows that discharge communication is often limited and the quality sub-optimal and associated with adverse health outcomes in 5 and 18 weeks after discharge.

OVERALL IMPRESSION

Overall this is an interesting and important topic that adds to existing knowledge, especially since it was conducted in a low and middle income country (India), where evidence is currently relatively scarce. However, there are some major and minor issues that the manuscript raised:

MAJOR ISSUES

Overall

o The title states: ‘patients’ attitudes and experiences of discharge communication..’, and this is also stated as the primary aim. Reading this, I was expecting a qualitative collection of this information, providing a comprehensive overview of patients’ positive and negative experiences with discharge communication, what their views and opinions regarding the concept are, etc. However, this aim only seems to be inventoried with one question in the questionnaire (from appendix 2: question 5) “Do you think it is important to get such a note at discharge? Why?”. For me, this question does not reflect the concepts ‘attitudes’ and ‘experiences’. With the other questions, in my opinion, the authors captured what patients remembered regarding the content of the discharge communication. Therefore, I would consider changing ‘attitudes’ and ‘experiences’ in the title and aims to ‘patient recall of verbal discharge communication’, as the authors already do in the results section. This would align the title and aim better with the actual outcomes, results, and conclusion.

o The primary aim is on the concept stated above ‘to investigate chronic NCD patient’ attitudes and experiences of discharge communication’. The secondary aim was ‘to explore the relationship between quality of discharge communication and health outcomes’. However, throughout the entire paper and especially in the discussion/conclusion section, the secondary aim receives way more attention, while the primary aim receives little to none, which raised some questions with me.

o Considering the changes in the past years with regards to healthcare policy (i.e. the use of electronic health information systems, like the authors state in the discussion), are these data and results (of dec 2014 – nov 2015) still relevant?

Introduction

o It could use some more explanation how the conclusion in the line 125-127 follows from the information in line 119-125.

o In the introduction there seems to be a lot of focus on communication between HCP’s and less on communication between HCP and the patient, while the latter is your study subject. In order to keep focus on the subject you are studying, I would focus more on the communication between HCP-patient, and/or explain the relation with between-HCP’s communication more.

Methods

o Were the patients recruited consecutively (24 hours a day?), what limitation was caused by the ‘available resources’?, line 195.

o In line 153-159 and figure 1. the authors provide an overview of the Indian public healthcare system, however I miss an explanation of how this relates and/or is of influence to the data and results. Potentially, this overview fits better in the discussion, if supplemented with the connection to this study.

o Line 164-185 gives a pretty comprehensive overview of the population, but this is also not discussed in the discussion: is the included sample representative for the population?

o What do the authors mean by ‘who had been admitted within 24 hour’ in line 197? Does that mean that patients were included within the first 24 hours of their admission?

o There is no explanation of whether patients with (known) cognitive decline or dementia were in- or excluded. This is important information, as the majority of included patients were aged 60 or older. Also, the study subject regards the patients’ recollection of the discharge communication, which of course could be highly influences by impaired cognition.

o There is no explanation of how illiterate patients filled out the questionnaire.

o In line 223-226 the authors state to have piloted the questionnaire: what was the conclusion of this pilot (was the questionnaire comprehensible and comprehensive, or did it need more adaption after the pilot?), and were the piloted cases included in the main study?

o Line 233: when looking at the datasheet in appendix 2; there seems to have been no data collected on patient experiences, but on patients’ recollection of information transfer.

o As line 245-250 addresses an important limitation of the study, this section should move to the discussion.

o Why are multivariable/multiple logistic regression performed? The authors mention one predictor (low quality of discharge communication) and not multiple, so I have some doubts whether this is the fitting statistical method.

o Unclear why the authors chose the Firth methods of logistic regression and whether this is the suitable method to use in this study. They state the reason was the small number of outcome measures, however, to my knowledge this is not a suitable reason to perform Firth method. The Firth method is used when the outcomes are very rare, and I am not sure whether these outcome measures are necessarily rare.

o Why was the predictor categorized dichotomously (low quality vs high quality), instead of 4 categories? Based on content of the categories I wonder whether all separate key-items should have the equal amount of weight. Does this mean that they all have the same amount of importance/impact? For example, shouldn’t appropriate medical information be more important than lifestyle advice?

Results

o Line 321-324 is about the understanding of patients, but how did the researchers test/judge whether patients/cares had a good understanding? This needs explanation in the methods section. Also, nowhere in the methods section is explained that carers could also be present an/or interviewed.

o The section ‘patient follow-up plans’ (line 350-357) was not elaborated on in the methods section, why was this measured and how does this add to answering the research questions?

o Line 408-411 (decreased odds of self-reported deterioration of NCD/s when patients’ had low quality verbal communication) : this shows an interesting, probably unexpected, finding: I miss a potential explanation for this.

o I am unsure whether the content of the sensitivity analysis really adds value.

o Line 459-462: the data does not support these claims, verbal leans in the other direction.

o Table 2: why were data of 120 patients missing regarding ‘patient/carer understanding of health condition post discharge’?

Discussion and conclusion

o Overall, I feel that many of the claims in the discussion are expressed too bold, which is not backed by the data and the results. For example: line 595-597: the authors have not shown significant associations with adverse health outcomes after 5 and 18 weeks; only death seemed to show a significant association with low quality documented discharge information.

o Line 527-541: I believe this section should move to the limitations as it is addressing a limitation of the study.

o I think the fact that there were 6 different researchers/coders might have caused some bias, so should be mentioned in the limitations.

o The next steps section is focusing on between HCP’s communication, I miss the link to between HCP-patient communication here.

MINOR

Introduction

o Line 109: add abbreviation LMICs, as the abbreviation is used in line 120.

o The South African study in line 116-118 is on discharge planning, not on communication. Explain why this is important information for this study.

Methods

o Line 171: abbreviations not explained in text

o Were interviews completed face-to-face or via telephone (line 231-231)?

Results

o Table 3: I believe the ‘*’ is at the wrong place?

o Line 389: supposed to be S6?

Discussion

o Line 523: compared to..

o Line 546-548: why? explain

Reviewer #3: Summary of the research

The aim of the study was to analyze patients’ experiences of discharge communication after presenting with a chronic disease in one of the three participating Indian hospitals. The second aim was to find out whether the quality of discharge communication could be correlated to certain adverse events.

Overall impression

- The paper is well written. The background literature and study rationale are clearly articulated. Conclusions are consistent with the evidence and arguments presented. The authors do address the main question posed.

- The collected data exceeds the minimal sample size needed as calculated by the authors.

- This is a multicenter study including three public hospitals in India (one rural secondary-care hospital, one peri-urban secondary-care hospital and one Urban tertiary-care hospital)

- Research ethics (e.g. participant consent, ethics approval) are addressed appropriately.

- The overall methods are clear, although I’m missing some details to be able to replicate the study (see: minor issues)

- Figures and tables are overall clearly presented

To improve the manuscript:

Minor issues

- P2 L40: low numbers should be spelled out > three instead of 3.

- P2 L43: Five instead of 5

- P2 L51 / L55: in general percentages should be preceded by whole numbers

- P8 L200: Do the authors have any information concerning the amount of potentially eligible patients and excluded patients? Where patients included 7 days a week? During office hours?

- P9 L227: Were patients interviewed or did they fill out the questionnaires themselves? Was this done in the same way for every time point?

- P9 L229: Why did the authors choose to interview patients at 5 and 18 weeks after discharge?

- P12 L288: Add percentage after whole numbers.

- P13 Table 1: ‘Time taken to reach hospital >1 hour’ should be < 1hour

- P13 Table 1: ‘Death’ meaning all-cause mortality? Of NCD specific?

- S4 Appendix: Almost all died patients were included at site 3, although this site only included half the amount of patients in comparison to site 1. Do the authors of the paper have an explanation for this?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

25 Feb 2020

Manuscript ID: PONE-D-19-22862

Manuscript Title: Investigating chronic disease patients’ attitudes and experiences of discharge communication in three hospitals in India

Dear Editors,

Thank you very much for reviewing our revised manuscript. We thank the reviewers for their generous feedback on our submission and have revised the manuscript accordingly.

Please find below a point-by-point response to reviewer’s comments - please note that the page and line numbers provided in each response correspond to the unmarked main document. We hope that you find our responses satisfactory and that the manuscript is now acceptable for publication.

With kind regards,

Dr. Semira Manaseki-Holland & Ms. Claire Humphries (BSc, MPH)

Institute of Applied Health Research

College of Medical and Dental Sciences

University of Birmingham

On behalf of all co-authors:

R. Lilford, P. Dorairaj, P. Gill, J. Panniyammakal, S. Singh, S. Greenfield, M. Price, and S. Jaganathan.

Reviewer 1

Major comments:

1. I am not sure whether the discovered associations between the (experienced) quality of the handover and health outcomes (especially all cause mortality) are that relevant, but these are a big part of the manuscript. As the authors also state in their discussion this relation has to be interpreted with caution as confounding bias (in my opinion) may play an important role. For instance, compliance to therapy has not been measured and I can only imagine the quality of handover being relevant in mortality if this is caused by an Adverse Event of the specific disease (instead of other causes). I think it's up to the editor to decide if this needs a change in the manuscript.

Thank you for your comments. Whilst the authors acknowledge the limitations of the regression analyses within the manuscript, the findings are considered relevant as they are supported by a body of literature indicating that suboptimal discharge practices can compromise continuity and safety of patient care.

Please note that amendments have been made to the discussion section of the manuscript to more suitably balance the coverage of results regarding the primary and secondary study objectives (see from line 475).

Minor comments:

2. Regarding the methods of this study I think it's a limitation not observing the verbal handover information. Now, the patients' recall is measured, which of course indicates how well patients understand the provided information; however, there might be some differences between what has been told and what is recalled by patients, which may have implications for quality improvement strategies. In addition, the teach-back method in my opinion could have been a better method instead of using multiple choice questions to measure recall.

Thank you for your comments. The lack of observation of verbal handover information has now been noted within the limitations section (see lines 573-575). Regarding delivery of questionnaires, they were administered by researchers in the form of an interview, where each question was read aloud and researchers then ticked the appropriate box/es for responses that corresponded to pre-defined answers (rather than answers being read aloud). Researchers were also able to write free-text notes in designated spaces for responses that did not correspond to pre-defined answers. This information has now been clarified in the methods section of the manuscript (see from line 253).

3. Concerning exclusion: are patients with cognitive impairment included as well (not stated they have been excluded)

Thank you for your query. Whilst there was no formal assessment of each patient’s cognitive abilities, researchers approached ward nurses to identify eligible patients. This process ensured that patients who were deemed too unwell to participate, due to severe physical and/or cognitive impairments, were excluded from the study.

In addition, all patients came to hospital with a carer (i.e. friend, relative etc.). Therefore, whilst patients predominantly provided questionnaire responses, carers were available to provide support for answering questions for any patients who required it.

The above information has now been clarified in the methods section of the manuscript (see from line 213).

4. You may consider to describe the pilot study a bit more in detail: how many patients were involved? Or did only researchers optimize the questionnaire? In that case, I think the study could have been optimized with pilot testing/cognitive interviewing of patients

Thank you for your queries. Prior to the commencement of data collection, a small pilot study was conducted in Kerala to test all data collection instruments. This was an iterative process conducted over three rounds. Field workers went out to study hospitals with the questionnaires and interviewed two patients/carers each. As well as asking patients/carers questions from the questionnaires/topic guides, they also asked for feedback regarding the clarity and contextual appropriateness of the materials. Once this was complete, all researchers convened with a supervisor to discuss the intended aim of each question and patient/carer responses. During these discussions, the wording within the materials was developed to improve clarity and contextual relevance. This information has now been clarified within the methods section of the manuscript (see from line 235).

5. Abbreviations do come up in text without previous being presented in (..). For example LMIC line 114, CHC/PHC/SHS line 171

Thank you for your comment- all abbreviations have now been presented in brackets prior to use throughout the manuscript.

6. Can you add a reference to the questionnaire used as a basis?

Thank you for your query. The previous research involving the original questionnaire is unpublished and PLOS One does not allow unpublished work to be cited within an manuscript, so we are not able to reference it. Instead, we have included it in the supplementary material (see S2 Appendix) for your reference.

Reviewer 2

Major comments:

1. The title states: ‘patients’ attitudes and experiences of discharge communication..’, and this is also stated as the primary aim. Reading this, I was expecting a qualitative collection of this information, providing a comprehensive overview of patients’ positive and negative experiences with discharge communication, what their views and opinions regarding the concept are, etc. However, this aim only seems to be inventoried with one question in the questionnaire (from appendix 2: question 5) “Do you think it is important to get such a note at discharge? Why?”. For me, this question does not reflect the concepts ‘attitudes’ and ‘experiences’. With the other questions, in my opinion, the authors captured what patients remembered regarding the content of the discharge communication. Therefore, I would consider changing ‘attitudes’ and ‘experiences’ in the title and aims to ‘patient recall of verbal discharge communication’, as the authors already do in the results section. This would align the title and aim better with the actual outcomes, results, and conclusion.

Thank you for your comments. The authors acknowledge that there is a predominant focus on recall and direct assessment of information transfer at discharge rather than an exploration of patients’ attitudes and experiences of discharge communication. This has now been more accurately reflected in the study title, as well as the aims and objectives.

2. The primary aim is on the concept stated above ‘to investigate chronic NCD patient’ attitudes and experiences of discharge communication’. The secondary aim was ‘to explore the relationship between quality of discharge communication and health outcomes’. However, throughout the entire paper and especially in the discussion/conclusion section, the secondary aim receives way more attention, while the primary aim receives little to none, which raised some questions with me.

Thank you for your comments. The authors disagree that the primary aim received little to no attention in the original manuscript; for example, lines 468-518 within the discussion were dedicated to findings regarding the primary objective and interpreting them in the context of the wider available literature. The sub-optimal nature of discharge communication and its consequences were also reported within the conclusion of the discussion section of the manuscript (lines 593-596). However, it is acknowledged that the secondary objective received a notable amount of attention and so amendments have been made to the discussion section of the manuscript to more suitably balance the coverage of findings in relation to the primary and secondary study objectives (see from line 477).

3. Considering the changes in the past years with regards to healthcare policy (i.e. the use of electronic health information systems, like the authors state in the discussion), are these data and results (of dec 2014 – nov 2015) still relevant?

Thank you for your query. Although healthcare policy regarding electronic health information systems has passed, only the state of Kerala has started to implement it. At present there is little central government money and most state governments are expected to fund their own systems implementation, which is a situation that is unlikely to be practical for most states in the near future (including Himachal Pradesh). In addition, within Kerala the electronic systems that are in place are currently used as more of a method for recording service utilisation and prescribing. They remain in their early stages and are far from the full integration that could effectively facilitate handover communication between HCPs. Therefore, the practices described in the study areas remain widely the same, with patient-held notes still in use throughout public healthcare and particularly at the point of hospital discharge. As a result, the authors remain firmly of the belief that the data and results described within the manuscript are still relevant.

Some further information clarifying current practices and progress of healthcare policy developments has been provided in the next steps of the discussion section of the manuscript (see from line 621).

4. It could use some more explanation how the conclusion in the line 125-127 follows from the information in line 119-125.

Thank you for your comment. Further information explaining how patients facilitate information transfer between healthcare providers within the study areas of India has now been provided in the introduction section of the manuscript (see from line 129).

5. In the introduction there seems to be a lot of focus on communication between HCP’s and less on communication between HCP and the patient, while the latter is your study subject. In order to keep focus on the subject you are studying, I would focus more on the communication between HCP-patient, and/or explain the relation with between-HCP’s communication more.

Thank you for your comments. The introduction section of the manuscript has been amended to more clearly demonstrate the significance and evidence regarding communication between HCPs and between HCPs and patients at the point of hospital discharge. In addition, further explanation has been provided as to why the study focussed on communication between HCPs and patients (see from line 90).

6. Were the patients recruited consecutively (24 hours a day?), what limitation was caused by the ‘available resources’?, line 195.

Thank you for your queries. Patients were recruited consecutively by trained social work graduate researchers six days per week (spread across all days of the week over the study period) between the hours of 8am and 6pm, as this is the window within which patients were typically discharged from study hospitals.

The wording “based on available resources” was simply used to demonstrate that as many patients were recruited as possible given the available human resources (i.e. six researchers).

Information regarding patient recruitment has now been clarified in the methods section of the manuscript (see from line 207).

7. In line 153-159 and figure 1. the authors provide an overview of the Indian public healthcare system, however I miss an explanation of how this relates and/or is of influence to the data and results. Potentially, this overview fits better in the discussion, if supplemented with the connection to this study.

Thank you for your comments. This was included within the main manuscript to provide general contextual information regarding the structure and functioning of public healthcare in India for readers who may not be familiar. It also gives an idea of the recommended coverage of each level of healthcare facility, which links to the next section regarding study populations that describes actual public primary care infrastructure in each area. The authors feel that this overview is better suited in the methods section, where the full text and figure can either be included in the manuscript or referred to via an appendix (depending on reviewer and/or editorial preference).

8. Line 164-185 gives a pretty comprehensive overview of the population, but this is also not discussed in the discussion: is the included sample representative for the population?

Thank you for your query. This information was again provided for contextual purposes and can either remain in full within the methods section or be referred to via an appendix (depending on reviewer and/or editorial preference). Given a lack of available data regarding the prevalence of chronic NCD patients and their demographic details in the study areas of India at the time of the study, it was not possible to make exact calculations to guarantee a fully representative sample. However, recruiting as many individuals as possible from healthcare facilities over a period of several months helped to ensure that the study sample was representative of chronic NCD patients visiting public hospitals within Himachal Pradesh and Kerala states, India. This was demonstrated by the recruitment of patients with a range of chronic NCDs, ages, education levels, employment statuses etc.

Some further information has been provided regarding representativeness of the study sample in the strengths of the discussion section of the manuscript (see lines 559-562).

9. What do the authors mean by ‘who had been admitted within 24 hour’ in line 197? Does that mean that patients were included within the first 24 hours of their admission?

Thank you for your query. Patients were included within the first 24 hours of their admission. This information has now been clarified in the methods section of the manuscript (see lines 210-211).

10. There is no explanation of whether patients with (known) cognitive decline or dementia were in- or excluded. This is important information, as the majority of included patients were aged 60 or older. Also, the study subject regards the patients’ recollection of the discharge communication, which of course could be highly influences by impaired cognition.

Thank you for your comments. Whilst there was no formal assessment of each patient’s cognitive abilities, researchers approached ward nurses to identify eligible patients. This process ensured that patients who were deemed too unwell to participate, due to severe physical and/or cognitive impairments, were excluded from the study. In addition, all patients came to hospital with a carer (i.e. friend, relative etc.). Therefore, whilst patients predominantly provided questionnaire responses, available carers could provide support for answering questions for those patients who required it. This information has now been clarified in the methods section of the manuscript (see from line 213).

11. There is no explanation of how illiterate patients filled out the questionnaire.

Thank you for your comment. Questionnaires were administered in the form of an interview and responses were completed by researchers, which meant that illiterate patients were not excluded from participation. This has now been clarified in the methods section of the manuscript (see from line 253).

12. In line 223-226 the authors state to have piloted the questionnaire: what was the conclusion of this pilot (was the questionnaire comprehensible and comprehensive, or did it need more adaption after the pilot?), and were the piloted cases included in the main study?

Thank you for your queries. Prior to the commencement of data collection, a small pilot study was conducted in Kerala to test all data collection instruments. This was an iterative process conducted over three rounds. Field workers went out to study hospitals with the questionnaires and interviewed two patients/carers each. As well as asking patients/carers questions from the questionnaires/topic guides, they also asked for feedback regarding the clarity and contextual appropriateness of the materials. Once this was complete, all researchers convened with a supervisor to discuss the intended aim of each question and patient/carer responses. During these discussions, the wording and structure of the questionnaire was developed to improve clarity and contextual relevance. After the three rounds of minor amendments, based on patient/carer and researcher feedback, the questionnaire was considered to be comprehensible and comprehensive. The piloted cases were not included in the main study. This information has now been clarified within the methods section of the manuscript (see from line 239).

13. Line 233: when looking at the datasheet in appendix 2; there seems to have been no data collected on patient experiences, but on patients’ recollection of information transfer.

Thank you for your comment. The authors acknowledge reviewer comments regarding the focus on information transfer rather than patients’ attitudes and experiences of discharge communication. This has now been more accurately reflected in the study title as well as the aims and objectives.

The authors also feel it is important to note that whilst the majority of data was based on patient recall, the data regarding documented information was extracted directly from medical notes (with patient permission) by researchers.

14. As line 245-250 addresses an important limitation of the study, this section should move to the discussion.

Thank you for your comment. This section has now been moved into the discussion section of the manuscript (see from line 577).

15. Why are multivariable/multiple logistic regression performed? The authors mention one predictor (low quality of discharge communication) and not multiple, so I have some doubts whether this is the fitting statistical method.

Thank you for your query. Multiple logistic regression was performed as the authors wished to try to adjust for potentially confounding variables. In the original manuscript, we stated from line 277: ‘Multivariable models adjusted for the following potential confounders: sex, age, education level, employment status, time taken to reach hospital, number of chronic NCDs and hospital site.’ Some further clarification has now been provided in the methods section of the manuscript (see lines 306-308).

16. Unclear why the authors chose the Firth methods of logistic regression and whether this is the suitable method to use in this study. They state the reason was the small number of outcome measures, however, to my knowledge this is not a suitable reason to perform Firth method. The Firth method is used when the outcomes are very rare, and I am not sure whether these outcome measures are necessarily rare.

Thank you for your comments. The authors disagree with the reviewer on this point. It is the (small) number of outcome events rather than the proportion (rarity) of the outcome that requires the Firth method.[1] In other words, the problem is not the rarity of the outcomes, but the possibility of a small number of cases on the rarer of the two outcomes.[2] The Firth method has become a relatively standard approach for analysing binary outcomes with smaller samples.[3]

References:

1. Firth D. Bias reduction of maximum likelihood estimates. Biometrika. 1993;80(1):27-38.

2. Allison P. Logistic regression for rare events. Statistical Horizons. 2012. Available from: https://statisticalhorizons.com/logistic-regression-for-rare-events

3. Puhr, R, Heinze, G, Nold, M, et al. Firth’s logistic regression with rare events: accurate effect estimates and predictions? Stat Med 2017; 36: 2302–231

17. Why was the predictor categorized dichotomously (low quality vs high quality), instead of 4 categories? Based on content of the categories I wonder whether all separate key-items should have the equal amount of weight. Does this mean that they all have the same amount of importance/impact? For example, shouldn’t appropriate medical information be more important than lifestyle advice?

Thank you for your queries. All four items of key information were selected based on common themes across relevant literature regarding critical details needed for facilitating post-discharge continuity and safety of care. The data were not included as a single four-category variable because multiple pieces of discharge information can be given at one time. Inclusion of each item of information as a separate dichotomous variable would be problematic. As well as leading to reduced power, the assumption of linearity on the logit scale would also be questionable. For example, the predicted change in log odds ratio for a patient due to receiving lifestyle advice, would be the same for a patient who received lifestyle advice only, as for a patient who received treatment/management information, medication information, lifestyle advice and follow-up instructions. This assumption would be untestable as there are insufficient data to include the six interaction terms that would be required to assess it.

18. Line 321-324 is about the understanding of patients, but how did the researchers test/judge whether patients/cares had a good understanding? This needs explanation in the methods section. Also, nowhere in the methods section is explained that carers could also be present an/or interviewed.

Thank you for your query and comments. Researcher judgements were made by asking patients/carers to verbally explain what they understood of their condition and post-discharge care requirements and then checking this information against their discharge notes, or if this was not available/did not contain enough information, a ward nurse ward who was aware of their condition and care requirements. This information has now been clarified in the methods section of the manuscript and it has been explained that whilst patients predominantly provided responses to questions, available carers were able to provide support with answers when necessary (see from line 257).

19. The section ‘patient follow-up plans’ (line 350-357) was not elaborated on in the methods section, why was this measured and how does this add to answering the research questions?

Thank you for your query. Patient attitudes and plans regarding follow-up care can be affected by healthcare provider advice and can ultimately affect continuity of care and patient outcomes. Therefore, information on patient-follow-up plans was collected to see how chronic disease patients planned to manage their ongoing care needs following the receipt of discharge consultation information. Ultimately, this information increased the authors’ understanding of whether patients were leaving hospital fully prepared to effectively manage their condition/s.

20. Line 408-411 (decreased odds of self-reported deterioration of NCD/s when patients’ had low quality verbal communication) : this shows an interesting, probably unexpected, finding: I miss a potential explanation for this.

Thank you for your comment. In the original manuscript the following potential explanation was offered for this finding in the discussion (lines 533 – 536): “For verbal information, it might be that those who were provide with less information at discharge were patients with less severe health issues (due to HCP time pressures/case prioritisation etc.) and were therefore less likely to report a deterioration later on”. The discussion section of the manuscript now contains a clearer and succinct summary of potential explanations for the findings from the regression analyses (see from line 536).

21. I am unsure whether the content of the sensitivity analysis really adds value.

Thank you for your comment. The authors take the view that the sensitivity analyses adds some value by addressing a limitation of the study. In order to make the manuscript more concise, more detailed information regarding the sensitivity analyses and results, which was previously part of the results section of the manuscript, has now been moved into an appendix (see supplementary material “S9 Appendix); This is referred to in the limitations of the discussion section of the manuscript (see from line 591).

22. Line 459-462: the data does not support these claims, verbal leans in the other direction.

Thank you for your comment. The sentence in question was referring to documented discharge information only and stated the following: “While no significant associations were found, all point estimates leaned in the direction of an increased likelihood of hospital readmission within five and eight weeks for those patients who received low-quality discharge notes”. The section of text regarding the sensitivity analyses has now been moved to an appendix and the description of results has been re-worded to improve clarity (see supplementary material “S9 Appendix”).

23. Table 2: why were data of 120 patients missing regarding ‘patient/carer understanding of health condition post discharge’?

Thank you for your query. A footnote below table 2 explains the following: “please note the larger number of missing responses to this question was due to a lack of available documented discharge information and/or ward nurses, which were required at the time of questioning for the researcher to verify patient/carer understanding of their condition and post-discharge care requirements”. Further information regarding how patient/carer understanding data was collected has now been clarified in the methods section of the manuscript (see from line 267).

24. Overall, I feel that many of the claims in the discussion are expressed too bold, which is not backed by the data and the results. For example: line 595-597: the authors have not shown significant associations with adverse health outcomes after 5 and 18 weeks; only death seemed to show a significant association with low quality documented discharge information.

Thank you for your comments. The discussion section of the manuscript has been amended to ensure the wording accurately represents the results and limitations of the study and it is maintained that the results of the regression analyses must be interpreted with caution (see from line 536).

25. Line 527-541: I believe this section should move to the limitations as it is addressing a limitation of the study.

Thank you for your comment. Further information regarding the limitations of the regression analyses has now been included in limitations of the discussion section of the manuscript (see from line 583).

26. I think the fact that there were 6 different researchers/coders might have caused some bias, so should be mentioned in the limitations.

Thank you for your comment. The possibility of researcher bias has now been listed as a potential limitation of the study in the discussion section of the manuscript (see lines 577-578).

27. The next steps section is focusing on between HCP’s communication, I miss the link to between HCP-patient communication here.

Thank you for your comment. Whilst the next steps section in the original manuscript did cover communication between HCPs and patients (e.g. in lines 567-570), the authors acknowledge that this could be made clearer. Therefore, to avoid confusion, the next steps within the discussion section of the manuscript has been amended to more clearly demonstrate suggestions for improving communication between HCPs and patients (see from line 598).

Minor comments:

28. Line 109: add abbreviation LMICs, as the abbreviation is used in line 120.

Thank you for your comment. The abbreviation has now been added (see line 112).

29. The South African study in line 116-118 is on discharge planning, not on communication. Explain why this is important information for this study.

Thank you for your comment. The South African study was described to demonstrate previous LMIC findings regarding the significance of discharge processes for patient outcomes. It was considered relevant as discharge planning is a process that involves the transfer of healthcare information between HCPs and between HCPs and patients to ensure coordination and continuity of care; In the literature it has been defined as: “An interdisciplinary approach to continuity of care; it is a process that includes identification, assessment, goal setting, planning, implementation, coordination, and evaluation and is the quality link between hospitals, community-based services, nongovernment organizations, and carers”[1].

Please note that the introduction section of the manuscript has been amended to improve relevance and clarity (see from line 81).

Reference:

1. Lin CJ, Cheng SJ, Shih SC, Chu CH, Tjung JJ. Discharge planning. International Journal of Gerontology. 2012 Dec 1;6(4):237-40.

30. Line 171: abbreviations not explained in text

Thank you for your comment. The abbreviations have now been explained in the text (see from line 182).

31. Were interviews completed face-to-face or via telephone (line 231-231)?

Thank you for your query. All interviews were completed in person (i.e. face-to-face). This has now been clarified in the methods section of the manuscript (see from line 253).

32. Table 3: I believe the ‘*’ is at the wrong place?

Thank you for your query. All symbols and corresponding footnotes for table 3 have now been checked for accuracy and corrected where necessary.

33. Line 389: supposed to be S6?

Thank you for your query. All references to supplementary materials throughout the manuscript have been checked and corrected where necessary.

34. Line 523: compared to.

Thank you for comment. This line has now been clarified within the discussion section of the manuscript (see from line 541).

35. Line 546-548: why? Explain (r.e. why follow-up data can be challenging to collect in LMIC settings)

Thank you for your query. Follow-up community-based data can be particularly challenging to collect and subsequently resource-intensive in LMIC settings such as India for a number of reasons. Some of these include:

• Difficulties in locating addresses due to a lack of comprehensive and up-to-date directories

• Inability to contact individuals to confirm attendance at follow-up visits due to a lack of phone ownership or limited network coverage

• Limited access to certain areas due to limited transportation infrastructure, challenging terrains and/or conflict/military presence.

This has now been clarified in the strengths and limitations in the discussion section of the manuscript (see from line 564).

Reviewer 3

Minor comments:

1. P2 L40: low numbers should be spelled out > three instead of 3.

Thank you for your comment. This has now been corrected in the abstract.

2. P2 L43: Five instead of 5

Thank you for your comment. This has now been corrected in the abstract.

3. P2 L51 / L55: in general percentages should be preceded by whole numbers

Thank you for your comment. This has now been corrected in the abstract.

4. P8 L200: Do the authors have any information concerning the amount of potentially eligible patients and excluded patients? Were patients included 7 days a week? During office hours?

Thank you for your queries. Due to the dynamic nature of the study settings, it was not possible to record inclusion/exclusion rates for participation. This has been noted as a limitation in the limitations within the discussion section of the manuscript (see lines 575-577).

Patient recruitment was completed by researchers 6 days per week (spread across all days of the week over the study period) between the hours of 8am and 6pm; these hours were selected as this is the period within which patients were typically discharged from study hospitals. Therefore, the authors suspect no bias in recruitment due to days of the week or hours of the day selected for recruitment. This information has now been clarified in the methods section of the manuscript (see from line 207).

5. P9 L227: Were patients interviewed or did they fill out the questionnaires themselves? Was this done in the same way for every time point?

Thank you for your queries. Questionnaires were administered by researchers in the form of an interview, where each question was read aloud and researchers then ticked the appropriate box/es for responses that corresponded to pre-defined answers (rather than answers being read aloud). Researchers were also able to write free-text notes in designated spaces for responses that did not correspond to pre-defined answers. This was done in the same way for every time point. This information has now been clarified in the methods section of the manuscript (see from line 253).

6. P9 L229: Why did the authors choose to interview patients at 5 and 18 weeks after discharge?

Thank you for your query. The follow-up questionnaires captured information on patient health-seeking behaviour. Therefore, patients were interviewed at five and eighteen-week follow-up due to the need for chronic NCD patients with ongoing health needs to have revisited healthcare providers within a month or at least within three months following hospital discharge.

7. P12 L288: Add percentage after whole numbers.

Thank you for your comment. All percentages within the text have the manuscript have now been rounded to whole numbers.

8. P13 Table 1: ‘Time taken to reach hospital >1 hour’ should be < 1hour

Thank you for your comment. This correction has been made in table 1.

9. P13 Table 1: ‘Death’ meaning all-cause mortality? Or NCD specific?

Thank you for your query. “Death” refers to all-cause mortality – this has now been clarified in table 1 and in the methods section of the manuscript (see lines 296-297).

10. S4 Appendix: Almost all died patients were included at site 3, although this site only included half the amount of patients in comparison to site 1. Do the authors of the paper have an explanation for this?

Thank you for your query. Site 1 and site 2 were hospitals situated in relatively non-deprived urban and peri-urban areas of Kerala. In these areas, patients have good levels of health literacy, there is a variety of healthcare providers available and the terrain and infrastructure makes travel reasonably easy for patients. Conversely, site 3 was a hospital situated in a rural and hilly area of Himachal Pradesh state, where patients are more socioeconomically deprived, generally less health literate and have longer distances to travel to a more limited range of healthcare providers. In addition, the climate is much cooler in this area so winters can be particularly harsh, with accompanying colds and flu that may have affected participants more than in Kerala. However, as we have no available data of the causes or circumstances of the deaths recorded in the study, we did not place too much emphasis on them and did not include the speculations above in the discussion – the authors can do this if reviewers /editor would like this inserted.

Submitted filename: Response to reviewers.docx

Click here for additional data file.

2 Mar 2020

DISCHARGE COMMUNICATION FOR CHRONIC DISEASE PATIENTS IN THREE HOSPITALS IN INDIA

PONE-D-19-22862R1

Dear Dr. Manaseki-Holland.

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Prof. Prabath Nanayakkara, MD, PhD, FRCP

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

26 Mar 2020

PONE-D-19-22862R1

Investigating discharge communication for chronic disease patients in three hospitals in India

Dear Dr. Manaseki-Holland:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. P W. B. Nanayakkara

Academic Editor

PLOS ONE