Pre hospital delay and its associated factors in acute myocardial infarction in a developing country.

BACKGROUND: Early revascularization and treatment is key to improving clinical outcomes and reducing mortality in acute myocardial infarction (AMI). In low- and middle-income countries such as Bangladesh, timely management of AMI is challenging, with pre-hospital delays playing a significant role. This study was designed to investigate pre-hospital delay and its associated factors among patients presenting with AMI in the capital city of Dhaka.
METHODS: This retrospective cohort study was conducted on 333 patients presenting with AMI over a 3-month period at two of the largest primary reperfusion-capable tertiary cardiac care centres in Dhaka. Of the total patients, 239(71.8%) were admitted in the National Institute of Cardiovascular Diseases, Dhaka and 94(28.2%) at Ibrahim Cardiac Hospital & Research Institute, Dhaka Data were collected from patients by semi-structured interview and hospital medical records. Pre-hospital delay (median and inter-quartile range) was calculated. Statistical significance was determined by Chi-square test. Multivariate logistic regression analysis was done to determine the independent predictors of pre-hospital delay.
RESULTS: The mean age of the respondents was 53.8±11.2 years. Two-thirds (67.6%) of the respondents were males. Median total pre-hospital delay was 11.5 (IQR-18.3) hours with median decision time from symptom onset to seeking medical care being 3.0 (IQR: 11.0) hours. Nearly half (48.9%) of patients presented to the hospital more than 12 hours after symptom onset. On multivariate logistic regression analysis, AMI patients with absence of typical chest pain [OR 5.21; (95% CI: 2.5-9.9)], diabetes [OR: 1.7 (95% CI: 1.0-2.9)], residing/staying > 30 km away from nearest hospital at the time of onset [OR: 4.3(95% CI = 2.3-7.2)] and belonged to lower and middle class [OR: 1.9(95% CI = 1.0-3.5)] were significantly associated with pre-hospital delays.
CONCLUSION: Acute myocardial infarction (AMI) patients with atypical chest pain, diabetes, staying far away from nearest hospital and belonged to lower and middle socioeconomic strata were significantly associated with pre-hospital delays. The findings could have immense implications for improvements about timely reaching of AMI patients to the hospital within the context of their sociodemographic status and geographic barriers of the city.

Background

Coronary artery disease (CAD) constitutes 15.9% of all deaths, making it the most common cause of death worldwide [1]. CAD is particularly prevalent in South Asia, with estimates from the Global Burden of Disease Study suggesting that the South Asian region will have more individuals with atherothrombotic cardiovascular disease than any other region in the days ahead [2]. The management of acute myocardial infarction (AMI) is extremely time-sensitive, with a delay in treatment of AMI being associated with increased mortality and morbidity [3]. The beneficial effect of fibrinolytic therapy was evident among patients at highest risk, including the elderly with proportional mortality reduction being significantly greater in patients treated within 2 h compared to those treated later (44% [95% CI 32, 53] vs 20% [15, 25]; p = 0.001) [4]. Early pharmacological or interventional reperfusion decreases mortality of ST-segment elevation myocardial infarction (STEMI). De Luca and associates showed 1 year mortality risk increased by 7.5% for each 30-minute delay in primary percutaneous coronary intervention (PCI) [3]. Furthermore, in non- ST-segment elevation myocardial infarction (NSTEMI) patients with a Global Registry of Acute Coronary Events (GRACE) score of more than 140, undergoing a coronary angiogram within 12 hours after admission was associated with lower risk of ischemic outcomes at 180 days [5].

Updated global cardiovascular society guidelines all recommend that reperfusion therapy is indicated in all patients with symptoms of ischemia of within a 12-hour duration and persistent ST-segment elevation for STEMI [6, 7]. An early invasive approach is also advocated in NSTEMI [8]. The 12 hour window for STEMI is particularly relevant in low and middle income countries (LMIC), where treatment delays [9, 10] due to lack of coordination between facilities, access to interventional cardiology facilities and catheterization laboratories contributed to poorer outcomes [11-14]. Several factors were found associated with increased time from symptom onset to treatment delays in AMI patients. Living in rural areas, hard road travel (poor road conditions or high traffic volume on the road) and lack of transport availability were reasons for prolonged pre-hospital treatment delays in several studies [15-18]. Mujtaba and colleagues [19] in a recent study in Pakistan observed that > 20% of the AMI patients attended the hospital after 6 hours of symptom onset with median pre-hospital time being 120 minutes (interquartile range: 229). The delay was more common among patients aged 41 to 65 years and among females. Factors that cause delayed presentation in hospital were misinterpretation, misdiagnosis, lack of transportation and financial constraint; of these, misdiagnosis act as a significant determinant of delay (p < 0.05). Financial constraint of the patients and the lack of a national insurance plan in many LMICs needs patients to pay out of pocket for cardiac reperfusion therapies which resulted delay in care or no access to care at all [20, 21].

There have been few studies exploring the pre-hospital delay of AMI presentations and its associated factors in Bangladesh. One study has recently been conducted in Northern city of Rajshahi [22] and another in Southern city of Chittagong [23]. On Online search, no study was found conducted in Dhaka, the capital city of Bangladesh. While several factors for delays were identified in these studies, it remains uncertain if they necessarily reflect those associated with patients presenting to specialized tertiary care hospitals in the capital city of Dhaka. The present study investigating the pre-hospital delay in patients with AMI was carried out in two specialized tertiary cardiac care centers in Dhaka to address this research gap in the previous studies.

Methodology

Study type and location

This retrospective cohort study was conducted at two of the largest primary reperfusion-capable cardiac hospitals in Dhaka, namely 1) National Institute of Cardiovascular Diseases (NICVD), Dhaka and 2) Ibrahim Cardiac Hospital and Research Institute (ICHRI), Dhaka. NICVD is the premier cardiac care centre of the government sector, serves patients at a significantly lower cost than centres in the private sector, and is visited by all classes of people. ICHRI is a tertiary cardiac care centre within the private sector, patronized by the more affluent sections of the society, as the cost of all services are borne by the patients themselves. Therefore, in choosing these two different hospitals, we ensured that we received patients from all strata of the society. The study hospitals were termed as index hospital in our study.

Sample size

The sample size for this study was determined by using the formula: where z = z value for 95% confidence interval, p = estimated prevalence of pre-hospital delay of more than 12 hours and d = precision time of error for the estimated prevalence. The proportion of patients with pre-hospital delay was derived from a previous study done in Chittagong, the Southern part of Bangladesh city was 62.7%. Thus, z = 1.96, p = 0.627 and d = 0.0627 (10% of p), calculated sample size was 228. Therefore, we included 333 patients presenting with AMI of which 239 (71.8%) were admitted in NICVD and 94 (28.2%) at ICHRI.

Inclusion and exclusion criteria

All consecutive patients who presented to ICHRI/ NICVD with AMI within the period of November 2019 to January 2020 and who consented to participate, were included. AMI was diagnosed according to criteria stipulated in the 4th universal definition of MI [24] and included patients with STEMI and NSTEMI. Patients who could not recall the events and those who received thrombolysis at a different hospital before presenting to ICHRI/NICVD were excluded.

Ethical clearance

The study protocol was approved by the Ethical Review Committees of both ICHRI and NICVD. Verbal consent was taken from each participant in presence of the ward doctors on duty and the participant had the right to withdraw from the study at any time during the study period. Confidentiality of participants was strictly maintained.

Dependent variable and independent variables

The total pre-hospital delay was the main dependent (outcome) variable. Total pre-hospital delay was defined as the time between the onset of symptoms of MI and time of arrival at the designated hospitals’ emergency room (ER), considering guideline recommended time cut-offs in recommendations for reperfusion therapy of STEMI [6, 7]. The pre-hospital delay was further divided into “Decision time” and “Time from decision to hospital ER”. “Decision time” was the time between the onset of patient’s symptoms and the time at which a decision was made to seek medical care. “Time from decision to hospital ER” is the time starting from a decision to go to hospital up to arrival at the hospital ER. This included time spent at any previous medical contacts and commute to the hospital. (Fig 1). Independent variables included age, sex, literacy, socioeconomic status, presence of typical chest pain, risk factors for CAD, prior history of CAD, first medical contact (FMC), distance from home and type of MI.

Fig 1

Working definitions for time intervals related to pre-hospital delay.

Working definitions

Socioeconomic status was categorized by monthly income: patients were asked about their occupation and approximate monthly income during the interview. A monthly income of more than 30,000 BDT (USD 357) were denoted as higher income, between 15,000 BDT (USD 178.5) and 30,000 BDT (USD 357) as middle income and below 15,000 BDT (USD 178.5) as low income.

Date collection, processing and statistical analysis

Data were collected by means of a pre-tested structured interview, supplemented by medical records. The patients were interviewed when they were clinically stable following treatment/revascularization and data were recorded on the questionnaire. Sociodemographic characteristics and history from symptom onset were obtained during interview, whereas clinical data were obtained from the patients’ hospital charts and electronic medical records. Data analysis was done by Statistical Package for the Social Sciences (SPSS) Version 25.0 (IBM). While continuous variables were expressed as median values with inter quartile range (IQR), categorical data were expressed as frequencies with percentages. Some quantitative variables like age and distance of the index hospital from home were dichotomized for univariate analysis. Univariate analysis was done using Chi-Square (χ2) test with Odds ratio (OR) to estimate the risk of having the outcome for a particular factor or characteristic. Variables found to be significantly associated with the outcome (pre-hospital) delay in univariate analysis were first subjected to Hosmer and Lemeshow Model-fit Test for Multivariate logistic (Binary logistic) regression analysis. After adjustment for confounding variables by Binary logistic regression analysis, the variables remained to be significantly associated with the outcome variable, were considered as independent predictors.

Results

Demographic characteristics, key risk factors, and clinical presentation

A total of 333 patients were studied, of whom two-thirds (67.6%) were males with male to female ratio being 2:1. Demographic and basic clinical characteristics are shown in Table 1. The mean age was 53.8±11.2 years (range 22–90 years). Over 60% presented with STEMI. Over three quarters of the total number of patients (76.9%) patients presented with typical chest pain as compared with 23.1% who presented with atypical symptoms. About 60% had diabetes, 60.1% had hypertension, 40.5% had a history of current/ previous smoking and 55% had family history of CAD. Nearly one-third (31.5%) of the patients were located within 10 km from hospital, 29.1% were at 10–50 km distance, while 39.3% were beyond 50 km distance at the onset of their symptoms (Range: 0.8–417 km). Around three-quarters (72.1%) received some form of preliminary anti-ischemic treatment from the referring clinic before arriving at a tertiary centre. Among them, 56.2% received dual anti-platelet therapy (DAPT) loading dose, 48.9% received sublingual nitroglycerine and 22.8% received low molecular weight heparin.

Table 1

Demographic and basic clinical characteristics.

Characteristics	Number	Frequency
Age
<35 years	12	3.6
35–50 years	101	30.3
51–70 years	194	58.3
Above 70 years	26	7.8
Sex
Male	225	67.6
Female	108	32.4
Employment status
Employed	159	47.7
Unemployed	174	52.3
Literacy
Illiterate	44	13.2
Up to Class 5	110	33.0
Up to Class 10	64	19.2
Up to Class 12	44	13.2
Undergraduate	55	16.5
Post-graduate	16	4.8
Socioeconomic Status
Lower income	113	33.9
Middle income	132	39.6
Higher income	88	26.4
MI symptom type
Typical chest pain	256	76.9
No chest pain	77	23.1
Self-perception and interpretation of symptoms by patient
Gastroenteric/Peptic Ulcer Disease	179	53.8
Muscular	33	9.9
Respiratory	26	7.8
Angina	77	23.1
None	18	5.4
Risk factors for CAD
Diabetes	199	59.8
Hypertension	200	60.1
Smoker/Tobacco	135	40.5
Family History of CAD	183	55.0
Prior history of CAD
None	214	64.3
Chronic Coronary Syndrome	105	31.5
Prior PCI	12	3.6
Prior CABG	2	0.6
Presentation chronology at tertiary hospital ER
First contact	66	19.8
Second contact	231	69.4
Third contact	36	10.8
Mode of transportation
Ambulance/EMS	146	43.8
Private transport	85	25.5
Public transport	102	30.6
Treatment received prior to arrival at tertiary hospital
None	93	27.9
Dual Antiplatelet Therapy loading	187	56.2
Sublingual Nitroglycerine	163	48.9
Low Molecular Weight Heparin	76	22.8

Overall, only 19.8% of patients presented directly to a tertiary cardiac centre as their FMC (first medical contact) (Table 1). Median total pre-hospital delay was 11.5 (IQR:18.3) hours with median decision time from symptom onset to seeking medical care being 3.0 (IQR:11.0) hours. About 70% of patients had presented to other hospital or clinic first, before eventually arriving at a tertiary centre for definitive care. Among the patients who have had previous medical contact before presenting to ICHRI/NICVD, 144 (53.9%) patients were referred by a general practitioner or community clinic from outside Dhaka city and 123 patients (46.1%) were referred from a different hospital within city areas.

Overall, only 19.8% of patients presented directly to a tertiary cardiac centre as their FMC (First Medical Contact) (Table 1). Median total pre-hospital delay was 11.5 (IQR: 18.3) hours with median decision time from symptom onset to seeking medical care being 3.0 (IQR: 11.0) hours. About 70% of the patients had presented to a single other hospital or clinic from onset of symptoms, before eventually arriving at a tertiary centre for definitive care. Among the patients who had previous medical contact before presenting to ICHRI/ NICVD, 144 (53.9%) patients were referred by a general practitioner or community clinic from outside Dhaka city and 123 patients (46.1%) were referred from a different hospital within city limits.

Pre-hospital delays

The distribution of the total pre hospital delay was skewed with a median of 11.5 (IQR-18.3) hours (Table 2). Of the total respondents, 163 patients (48.9%) presented to the hospital more than 12 hours after symptoms onset and the rest 170 (51.1%) within 12 hours of the symptoms. Median decision time from symptom onset to seeking medical care was 3.0 (IQR-11.0) hours. Median time from decision to arrival at hospital ER was 5.0 (IQR-8.0) hours (Figs 2–4). The time from decision to arrival at hospital ER was shorter for those that directly presented to the tertiary hospitals as compared with patients who were referred through a general practitioner (GP), community clinic or a district hospital [2.0 (IQR-2.0) vs. 6.0 (IQR-9.5) hours respectively.

Table 2

Total pre-hospital delay, decision time and time from decision to reach hospital ER.

Statistics of pre-hospital delay	Decision Time (hours)	Time from home to NICVD/ICHRI (hours)	Total time to receive treatment (hours)
Mean (SD)	10.7(15.7)	9.4(14.9)	20.3(22.4)
Median (IQR)	3(11.0)	5(8.0)	11.5(18.3)
Skewness	2.2	3.3	1.9

*IQR = Interquartile Range.

Fig 2

Histogram showing distribution of decision time with right skewness.

Fig 4

Histogram showing distribution of time to receive treatment with right skewness.

Factors associated with pre hospital delay

Table 3 shows the factors associated with outcome (pre hospital delay) between those presenting within and beyond 12 hours of symptom onset, with odds ratios (OR) in univariate analysis. Pre-hospital delay was significantly longer in females than males [OR:1.9; (95% CI: 1.2–3.1)], although age did not influence on pre-hospital delay. Respondents from lower and middle-income households were at 2.6 (95% CI: 1.5–4.1) times higher risk of having pre-hospital delay. Patients with level of education below 10th grade tend to be associated with pre-hospital delay with odds of having the condition being 2.3 (95% CI: 1.4–3.5) times higher. Diabetics were more prone to have pre-hospital delay than their non-diabetic counterparts with risk of having the condition 2.1 (95% CI: 1.3–3.3) times greater in the former group than that in the latter group. Patients presenting with atypical chest pain had significantly greater delays than those with typical chest pain with odds of having delay being much higher in the former cohort [5.8 (95% CI = 3.1–10.5)]. Significantly longer pre-hospital delays were also noted in patients located beyond 30 km from the hospital at symptom onset [3.9 (95% CI = 2.5–6.1)] (Table 3).

Table 3

Factors associated with pre-hospital delay.

Demographic, clinical characteristics & geographic barriers	Delay		Odds ratio	*P-value
	> 12 hrs (n = 163)	≤ 12 hrs (n = 170)	95 CI of OR
Sex
Female	65(39.9)	43(25.3)	1.9(1.2–3.1)	0.004
Male	98(60.1)	127(74.7)
Age years
>50	112(68.7)	108(63.5)	1.3(0.8–1.9)	0.318
≤50	51(31.3)	62(36.5)
Socioeconomic Status
Lower & middle class	134(82.2)	111(65.3)	2.6(1.5–4.1)	<0.001
Upper class	29(17.8)	59(34.7)
Level of education
Below 10th grade	92(56.4)	62(36.5)	2.3(1.4–3.5)	<0.001
10th grade and above	71(43.6)	108(63.5)
Diabetes
Present	112(68.7)	87(51.2)	2.1(1.3–3.3)	0.001
Absent	51(31.3)	83(48.8)
Hypertension
Present	91(55.8)	109(64.1)	0.7(0.5–1.1)	0.123
Absent	72(44.2)	61(35.9)
Smoker/Tobacco use
Yes	71(43.6)	64(37.6)	1.3(0.8–1.9)	0.272
No	92(56.4)	106(62.4)
Family History IHD
Present	97(59.5)	86(50.6)	1.4(0.9–2.2)	0.102
Absent	66(40.5)	84(49.4)
MI symptom type
No Chest Pain Atypical	61(37.4)	16 (9.4)	5.8(3.1–10.5)	<0.001
Chest Pain Typical	102(62.6)	154 (90.6)
Previous experience of IHD, chronic stable angina, PCI or CABG
Present	61(37.4)	58(34.1)	1.6(0.7–1.8)	0.529
Absent	102(62.6)	112(65.9)
Presentation at NICVD/ICHRI
Direct presentation to ICHRI/NICVD	129(80.1)	136(80.0)	1.0(0.6–1.7)	0.977
Referred from other centers	32(19.9)	34(20.0)
Transportation used
Private/Public Transport	96(58.9)	91(53.5)	1.2(0.8–1.9)	0.324
EMS/Ambulance	67(41.1)	79(46.5)
Distance from NICVD/ICHRI
> 30 km	108(66.3)	57(33.5)	3.9(2.5–6.1)	<0.001
≤ 30 km	55(33.7)	113(66.5)

*Chi-squared χ2 Test was done analyze the data; figures in the parentheses denote corresponding.

Analyses of association between diabetes and chest symptoms demonstrated that majority of the patients with atypical chest pain (84.4%) had diabetes as compared 52.3% of the patients with typical chest symptoms with risk of having atypical chest pain in diabetics being 4.9 (95% CI: 2.5–9.5) times higher than those without diabetics (Table 4).

Table 4

Association of diabetes with chest symptoms.

Diabetes	Chest Pain		Odds ratio 95 CI of OR	*P-value
	Atypical n = 77	Typical n = 256
Present	65 (84.4%)	134 (52.3%)	4.9(2.5–9.5)	< 0.001
Absent	12 (15.6%)	122 (47.7%)

Table 5 demonstrates the binary logistic regression analysis of odds ratios for factors associated with pre-hospital delay. The variables found statistically significant in univariate analysis (sex, socioeconomic status, level of education, diabetes mellitus, presence of typical chest pain and distance from the hospital) were all directly entered into the model-fit test first for binary logistic regression analysis. Hosmer and Lemeshow goodness-of-fit test demonstrated that the model was a good-fit-model which could correctly predict the outcome of interest (pre-hospital delay) in 69.3% of the patients (p = 0.081) with overall correct prediction capability of the model being 80.0%. Regression analysis revealed that patients with absence of typical chest pain, residing/staying > 30 km from the index hospital at time of symptom onset, having diabetes and belonged to lower- and middle- socio-economic class were independently associated with pre-hospital delay with odds of having the condition in these cohorts being 4.9 (95% CI: 2.5–9.9), 4.3 (95% CI: 2.3–7.2), 1.7 (95% CI: 1.0–2.9) and 1.9 (95% CI: 1.0–3.5) respectively.

Table 5

Factors associated with pre-hospital delay in multivariate logistic regression.

Factors	Univariate Analysis		Multivariate Analysis
	OR (95% CI of OR)	p-value	OR (95% CI of OR)	p-value
Sex	1.9(1.2–3.1)	0.004	1.7(0.9–1.7)	0.099
FemaleR vs. Male
Socioeconomic Status	2.6(1.5–4.1)	<0.001	1.9(1.0–3.5)	0.036
Lower & Middle incomeR vs. Higher income
Literacy Status	2.3(1.4–3.5)	<0.001	1.6(0.9–2.9)	0.084
Below 10th gradeR vs. Above 10th grade
DM	2.1(1.3–3.3)	0.001	1.7(1.0–2.9)	0.043
PresentR vs Absent
MI symptom type	5.8(3.1–10.5)	<0.001	4.9(2.5–9.9)	<0.001
No chest painR vs Chest pain
Distance from hospital	3.9(2.5–6.1)	<0.001	4.3(2.3–7.2)	<0.001
>30 kmR vs ≤30 km

R = Reference case.

Discussion

The main aim of this study was to identify predictors of pre-hospital delay for AMI patients in Dhaka, the capital of Bangladesh. We conducted the study at two hospitals: NICVD, a state-run hospital, provides services to a huge volume of patients at a negligible cost and ICHRI is a private hospital requiring patients to finance the whole course of their treatment. Combined, these two hospitals represent a patient population covering all socioeconomic strata of the city. Being two of the largest cardiac tertiary hospitals in the country, this study includes patients coming from various cities of Bangladesh, reflecting the contemporary status of the city’s cardiac care. There is scarce data on trends and factors responsible for pre-hospital delays in seeking cardiovascular care in Dhaka. This study is unique from prior studies investigating pre-hospital delays in the country for many reasons: 1) the capital city is home to a large population (21,741,000) [25] of various socioeconomic statuses, some of whom can only afford cardiac care in a state hospital, while more affluent people opt for self-financed care in private hospitals. Besides, the capital city receives patients not only from within the city boundaries but across the country, for some of whom definitive cardiac care is only first administered when they reach a tertiary centre.

The median pre hospital delay in our study population was 11.5 (IQR-18.3) hours and the median decision time was 3.0 (IQR-11.0) hours. While the “decision time” is almost entirely attributed to patient factors, the time taken from decision to arrival at NICVD/ICHRI depends on various factors such as referral time, mode of transport, traffic, etc. A previous study by Rafi, et al [22] in Northern Bangladesh revealed a median pre hospital delay of 9 (IQR-13) hours with a median decision time of 2.0 (IQR-1.0) hours. Compared with our study, factors contributing to these slightly lower pre-hospital delays in their study could be attributed to the fact that it was conducted in Northern Bangladesh, where traffic is not as problematic as in the capital city of Dhaka. Furthermore, the hospital at which the study was conducted was a tertiary care centre directly serving the catchment area, which is smaller than that of the present study, and with fewer referrals from across the country, in comparison to those in our study.

Analysis of the factors contributing to pre-hospital delay demonstrated that patients with atypical chest symptoms, residing/staying > 30 km from the index hospital at symptom onset, being belonged to lower and middle income socio-economic class and diabetes were independent predictors of pre-hospital delay with odds of having the condition in these cohorts being 4.9 (95% CI: 2.5–9.8), 4.5 (95% CI: 2.3–7.5), 1.9 (95% CI: 1.0–3.5) and 1.7 (95% CI: 1.0–2.9) respectively. Similar to our study, Rafi, et al [22] also found diabetes, lower socio-economic conditions, absence of typical chest pain and greater distance between site of symptoms onset to hospital to be significantly associated with pre-hospital delay. Similar observations were reported in a study by Das et al [23]. They found a mean delay of 6.8 ± 3 hours for those presenting early (as < 12 hours), and a mean delay of 37.8 ± 25.1 hours for late presenters. In addition to distance and diabetes, they found greater delays in those with elderly and those who misinterpreted symptoms for peptic ulcer disease (PUD). Our study results consistent with those from neighbouring LMIC, such as India [15, 26], who also have high pre-hospital delays, and contrasts with high income countries, such as Sweden [27], Australia [28], United States of America [29] and Poland [30] showing a significantly lower pre-hospital delay in patients with AMI.

There are paradoxical finding in literature regarding sex differentials of pre hospital delay. A study conducted in Iran by Farshidi et al [31] and another from Chittagong, Bangladesh [23] showed no significant differences in pre hospital delay between male and female AMI patients. Alternately, in a multinational registry reported by Bugiardini et al [32] and another study conducted in Sweden by Lawesson et al [33] showed that women had significantly longer pre-hospital delays in AMI presentation. In our study, while univariate analysis showed sex as a significant factor, it emerged non-significant on multivariate analysis, indicating sex as a potential confounder for the pre-hospital delay.

As Bangladesh is a developing country, only a negligible percentage of the population has adequate health insurance coverage. In major private institutions offering standard cardiac care, all treatments received by a patient are entirely self-financed, including access to emergency medical service (EMS), which possibly factors in the delays in decision time in patients with AMI, which is particularly reflected in the significant differences in pre-hospital delay times across different socioeconomic statuses. Patients with a lower income of less than 30,000 BDT (~357 USD) per month may delay the decision to seek help due to financial reasons for which they have a significantly high pre hospital delay, which is consistent with another study [34]. This is further compounded by possibly a lack of health-seeking behavior owing to a lack of awareness of the time-sensitive nature of cardiac care, and also a dependency on other family members to make key decisions, which largely stems from a general cultural tendency to involve multiple parties in decision-making. Concurrently, patients who have studied beyond the 10th grade showed a significantly lower decision time, possibly resulting from increased knowledge and awareness. Lesser education level has also shown to increase the delay in an Australian study as well [28]. Our study showed that those with diabetes mellitus had the significantly longer decision times. Diabetes Mellitus is not only a major risk factor of coronary artery disease, it also results in silent MI [35] and more atypical symptoms, all of which could contribute to delays in recognizing symptoms, and hence delays in presentation. A large percentage (53.8%) of the patients in our study attributed their symptoms to PUD or acid reflux. Patients who presented with chest pain presented to the hospital relatively earlier than patients who didn’t have chest pain as their symptom. This is concurrent with most studies [22, 27].

In contrast to studies in high income countries [36], the mode of transportation to hospital had no significant effect on the pre hospital delay in our study. In high income countries, it has been shown that EMS-transported patients had significantly shorter delays in symptom onset to arrival time in hospitals. However, within Dhaka city limits and indeed in neighboring suburbs, persistent traffic and the lack of emergency lanes preclude any benefit that may be derived from ambulance transportation; furthermore, an ambulance dispatch might even result in further delays considering a lack of organized EMS infrastructure and the time factored in to reach the patient. Additionally, few ambulances are equipped with trained staff that can administer DAPT, LMWH or thrombolytics with appropriate monitoring; as such many patients opt for personal vehicles and more easily available taxi services to reach hospitals. The time delay from FMC also depends on the distance from patient location to FMC: two-thirds (66.3%) of the patients who had delayed presentation reached the index hospitals from a location beyond 30 km as compared to (33.5%) of the patients who reached the hospital within 12 hours.

Although we did not assess clinical outcomes, a delayed presentation of MI is associated with an increased in-hospital mortality according to several studies [37, 38]. Reperfusion therapy is the key strategy to reduce mortality in AMI but its benefit is time dependent, and a 12-hour cut-off is particularly important in case of STEMI [6, 39]. Thus, reduction in the pre hospital delay will reduce mortality in patients with AMI.

This is a study that reflects the contemporary and realistic trends of pre-hospital delay in seeking cardiac care in a LMIC setting. It also highlights the multiple underlying issues of a healthcare system that need to be addressed in order to improve parameters and ensure timely cardiac care. The absence of a competent and well-trained EMS system remains a huge problem compromising timely cardiac care. This is further compounded by the absence of a functional referral system that is structured in a manner that at the very least ensures preliminary care and reperfusion for patients, prior to being transferred for tertiary care.

This study also highlights the variations in FMC sought and tendency of the general public to not adhere to referral systems. Whether this is a general lack of trust in the referral hierarchy, possibly stemming from an absence of fully-equipped treatment for AMI in many non-tertiary centres, or a matter of convenience in directly accessing a specialist, is a matter of debate. Alternatively, the habits of visiting local GPs instead of presenting to hospital ER, reflects a need for public health messaging and education of the general public on appropriate self-presentations for acute medical care. This can be done by locally and socially appropriate, community-targeted campaigns and interventions to increase the knowledge and awareness about the disease and its symptoms, and encourage a healthcare-seeking behavior among the general population. In the currently existing set-up, there is an additional need of adequate training of GPs and non-specialist doctors to promptly administer essential medicines in the management of AMI. Progressive modifications of healthcare infrastructure, with particular focus in improving emergency medical care, training and developing a more competent cohort of paramedics and ambulances, and implementing a proper and well-adhered to referral system, are important factors that need to be considered in decreasing pre-hospital delays in cardiac care.

Strengths and limitations

Patients’ symptoms, diagnoses and other clinical information were obtained from electronic medical records that reduced the risk of errors. Although this study was conducted in two of the largest cardiac hospitals in the country with a large turn over and covered a large portion of urban and peri-urban population, the findings do not necessarily represent those of entire reference population of the country. As > 12-hour delay is a sensitive issue for AMI patients, we dichotomized the delay in ≤ 12 and > 12 hours to see what factors could contribute to > 12 hours delay. However, dichotomizing the hours of delay and doing logistic regression have lost some information on understanding quantitative hours of delay. It could be better modeled with an over dispersed Poisson regression or negative binomial regression analysis. Besides, the study also did not involve the patients who could not make it to the hospital or died within a few hours of arrival. So, caution should be exercised to generalize the findings to reference population.

Conclusion

This present study investigating pre-hospital delays in the presentations of AMI patients in the Bangladeshi capital city of Dhaka revealed that the presence of diabetes, lower socioeconomic status, absence of chest pain and a distance from hospital greater than 30 km were significantly associated with increased pre-hospital delays in presentation. Educating people, particularly of low socioeconomic class having diabetes and residing far away from the cardiac care hospitals about AMI symptoms and importance of early presentation to hospital with concurrent improvements of the existing referral system and EMS will go a long way to reduce delays.

Transfer Alert

This paper was transferred from another journal. As a result, its full editorial history (including decision letters, peer reviews and author responses) may not be present.

12 May 2021

PONE-D-21-11112

Pre-hospital delay and its associated factors in Acute Myocardial Infarction in a developing country

PLOS ONE

Dear Dr. Chowdhury,

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. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Although all reviewers and editors found that this manuscript has merit and addresses a significant clinical problem, the reviewers raise some important issues. These comments include data presentation and statistical analysis (reviewer #1 and #2), discussion to be improved (reviewer #1 and #2), insufficient data on patients’ background, confounders, and quantitative hours of delay (reviewer #2), and sex differences (reviewer#2). In addition, among 333 patients, the number of patients who visited NICVD or ICHRI should be provided separately.

Please submit your revised manuscript by Jun 26 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're 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.

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). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Michinari Nakamura, MD

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

and

2. Please provide additional details regarding participant consent.

In the ethics statement in the Methods and online submission information, please ensure that you have specified what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed).

If your study included minors, state whether you obtained consent from parents or guardians.

If the need for consent was waived by the ethics committee, please include this information.

[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

**********

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

Reviewer #1: Yes

Reviewer #2: No

**********

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

**********

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

**********

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: OVERALL SUMMARY

This study presents conclusions for a cross sectional study conduct in a low-income country examining pre-hospital delays for patients seen for cardiovascular disease outcomes adjusting for potential confounders. The work is insightful and thought provoking. See below for comments from my review.

INTRODUCTION

The authors do not discussion findings from previous studies examining pre-hospital delay for cardiovascular outcomes.

Paragraphs 2 and 3 can be combined as they are talking about the same thought. After shortening the combined paragraph more can be added which presents results from actual scientific studies as opposed to just medial guidelines.

The introduction is not sufficient as it stands. The authors need to use past studies to make a case for the issue they are studying and point out the gaps in previous research that their research addresses.

METHODS

The authors need to define mid to lower socio-economic class populations. What income level is this? Can be different depending on country. This first paragraph on study location can be shortened. Its adequate to talk about the medical facility, its resources, and the population being studied. It would also be good to add where geographically these facilities are located and what is the median income/ other economic variables associated with the region. Have other studies also categorized this area as underserved? Should include a citation if so.

Bolding and quotations marks are not needed ** See author formatting guidelines.

Images should be included at the end of the text and follow journal formatting guidelines. ** See author guidelines.

RESULTS

Results from the 95% confidence interval should be presented in text with odds ratios in place of the p-values. P-values can be presented in the table alongside the 95% CI and OR estimates. However, it is more appropriate to present confidence intervals when discussing the odd ratios.

DISCUSSION

How do the demographics for your study compare with the demographics living in the study area? For example, a majority of patients in this study were 51-70 years of age (no doubt due to the condition of interest). It would be interesting to know if the area served by the medical facility is located in a generally older population setting.

The discussion was well thought out and included comparison to other studies.

Reviewer #2: Pre-hospital delay and its associated factors in Acute Myocardial Infarction in a developing country

This paper complements existing literature on AMI care in LMICs. The main logistic regression analysis could be strengthened with a stronger theory-based rationale for the variables used. The authors seem to choose the variables for the model empirically, rather than from a theory of what is most important. They describe the findings after the fact — e.g., reasons for diabetes associated with more likelihood of >12 hour delay because diabetes is associated with low-symptom AMI — but they should test these ideas within the data that they have. These additional analyses don’t need their own tables, but they can be reported within the text.

The analysis should as much as possible derive from which factors are most influential for policy and practice, such as identifying which populations to be targeted.

The hours of delay as a quantity is also important due to effect of each increment of delay on outcomes: e.g., authors say each 30 minute delay before PCI associated with greater 1-year mortality. dichotomizing the hours of delay and doing logistic regression loses information, and understanding quantitative hours is also important. In addition to the logistic regression, hours of delay is a count variable, and it could be modeled with an over dispersed Poisson or negative binomial regression.

The authors very clearly justify the 12 hour dichotomized variable. This dichotomized variable isn’t a rare event, so logistic regression give odds ratios that are much larger than the relative risks. A model such as Poisson with robust standard errors or negative binomial for this binary variable would give prevalence ratios, which can be interpreted as relative risks, and will be smaller than the odds ratios.

Women have different presentations of AMI, different social roles, and table 3 indicates that the median time to presentation is 21 hours. That’s astounding. Consider separating analysis into men and women, to see whether the associations differ.

The authors describe this study as cross-sectional. However, the authors followed the patients by medical records in addition to the survey, so it would be reasonable to describe this as a retrospective cohort study. The investigators’ design is actually a very usual retrospective cohort study.

The power is well-justified for a bivariate analysis. However, the authors do a multivariate regression analysis. The paper would be stronger with some type of sample size justification. Consider Gelman and Carlin’s Type S and Type M errors, at least to estimate probability that the estimates are in wrong direction.

Sentences shouldn’t begin with a number.

Table 1 includes prior PCI and CABG, but this section seems like it should have also prior AMI as complementary information. — presumably this information was recorded and excluded as oversight?

The authors group smoking/tobacco into just one variable. Mode of ingesting tobacco may be important to report separately (e.g., bidi), and also betel/areca nut, if these were asked on survey.

Table 2 gives just a few numbers and would be much more clear as a data display because clearly there’s a long tail on this distribution on time to arrive to hospital, and it’s very right-skewed. The information could be displayed in 4 histograms, for instance.

Tables should start on their own page and not run to the next.

Table 3 has a typo: under total column, percent of women is listed as 62.4%.

Age as dichotomized variable loses so much information. 3-5 groups may be more reasonable for age so the authors have a chance to detect an age effect.

The p-value in table 3 is strictly to test the dichotomized version of the variable. However, a test of the quantitative variable would also be useful. It doesn’t need its own column: test the medians using Wilcoxon and use asterix or bold to indicate significance. I suspect 4 hour median difference by hypertension status is significant.

The authors find in table 3 that in general people with risk factors take longer to get to the hospital: people with hypertension, diabetes, family history, and smokers. Why is this? Is this association due to socioeconomic status? Age? The authors can evaluate this question with their data.

The authors put a lot of variables into their logistic regression model that have multicollinearity, such as education and income. Putting many variables into a model and seeing what is significant doesn’t actually find out what is most important. The limited sample size makes this difficult, but some variables are so important that they could be stratified on. For instance, typical vs atypical symptoms, which is the largest effect. The variables associated with longer time for the atypical symptoms may differ from the variables important for people with typical symptoms. Distance from hospital is also very important. Other variables may be more appropriate to include as numbers, such as distance from hospital and income and years of education.

The regression has diagnosis as a predictor of delay, but diagnosis happened after the delay (after going to the hospital, that is) rather than before, so it’s not logical to put it into the regression. If there is a variable that is associated with the NSTEMI vs STEMI diagnosis that is apparent prior to going to the hospital, that would be more reasonable. In discussion authors mention less severe symptoms and say they can’t test. However, the symptom related variables collected in the survey may be associated with these.

The implications of this research for policy and practice should be thought about before choosing variables. For instance, if the question is which populations should be targeted for messaging about AMI symptoms, then the variables should correspond to demographic variables and also include interaction terms: e.g., between income and gender because low-income women may have longer delays than low-income men and than women overall.

“Diabetes Mellitus is not only a major risk factor of coronary artery disease, it also results in silent MI (25) and more atypical symptoms, all of which could contribute to delays in recognizing symptoms, and hence delays in presentation.” — this is interesting, and it’s a hypothesis that could be tested with the symptom data that was collected.

**********

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

[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.]

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 PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

9 Sep 2021

Thanks a lot to both the reviewers for an excellent feedback. Apart from making my manuscript more technically sound, I also learnt a great deal from these comments. I apologize for submitting the revision this late because there has been a huge surge in COVID cases in my country the last few months, so the coauthors and I were overwhelmed with our workload in our hospitals and also had to maintain a social bubble so could not sit and work on this together. I am addressing each point of the reviewers below.

Reviewer#1

I am very grateful to you for such a wonderful feedback. I corrected my manuscript accordingly and addressing each point below

Comment:

The authors do not discussion findings from previous studies examining pre-hospital delay for cardiovascular outcomes.

Paragraphs 2 and 3 can be combined as they are talking about the same thought. After shortening the combined paragraph more can be added which presents results from actual scientific studies as opposed to just medial guidelines.

The introduction is not sufficient as it stands. The authors need to use past studies to make a case for the issue they are studying and point out the gaps in previous research that their research addresses.

Reply: Thank you for pointing it out. Paragraph 2 and 3 of “Introduction” section have been merged and shortened. Relevant findings from other studies have also been added to make the study stand to reason. The research gap has also been highlighted.

Comment: The authors need to define mid to lower socio-economic class populations. What income level is this? Can be different depending on country. This first paragraph on study location can be shortened. Its adequate to talk about the medical facility, its resources, and the population being studied. It would also be good to add where geographically these facilities are located and what is the median income/ other economic variables associated with the region. Have other studies also categorized this area as underserved? Should include a citation if so.

Reply: We have added the income level in both Bangladeshi currency as well as US dollars. The first location has also been added and a brief insight has been given regarding the resources and population being studies. Both the hospitals are location in Dhaka, the capital city of Bangladesh. There is not enough data on the median income or other economic variables associated with this region. Moreover, both the hospitals serve patients from all parts of the country so the catchment area is very large.

Comment: Bolding and quotations marks are not needed ** See author formatting guidelines.

Images should be included at the end of the text and follow journal formatting guidelines. ** See author guidelines.

Reply: Pardon me for the silly mistake. I have now corrected the format according to guidelines

Comment: Results from the 95% confidence interval should be presented in text with odds ratios in place of the p-values. P-values can be presented in the table alongside the 95% CI and OR estimates. However, it is more appropriate to present confidence intervals when discussing the odd ratios.

Reply: Thanks for enlightening me regarding this. I have corrected the results accordingly. Odds ratios with its 95% confidence interval has been presented in both text and table with p-values alongside the OR(95% CI).

Comment: How do the demographics for your study compare with the demographics living in the study area? For example, a majority of patients in this study were 51-70 years of age (no doubt due to the condition of interest). It would be interesting to know if the area served by the medical facility is located in a generally older population setting.

Reply: This is a very important point. Unfortunately we don’t have enough data on age distribution in the population of concern.

Reviewer#2

Thank you very much to you for such details comments. Not only did all these comments make my manuscript as meticulous as I could, it also taught me a great deal as a junior doctor/researcher. Each comment you made was very valid and resonated well with our coauthors. Unfortunately, being from a low-middle income country, we are not too acquainted with some of the statistical methods. We revised the manuscript as much as we could within our capabilities. Thank you once again.

Comment: The hours of delay as a quantity is also important due to effect of each increment of delay on outcomes: e.g., authors say each 30 minute delay before PCI associated with greater 1-year mortality. dichotomizing the hours of delay and doing logistic regression loses information, and understanding quantitative hours is also important. In addition to the logistic regression, hours of delay is a count variable, and it could be modeled with an over dispersed Poisson or negative binomial regression.

The authors very clearly justify the 12 hour dichotomized variable. This dichotomized variable isn’t a rare event, so logistic regression give odds ratios that are much larger than the relative risks. A model such as Poisson with robust standard errors or negative binomial for this binary variable would give prevalence ratios, which can be interpreted as relative risks, and will be smaller than the odds ratios.

Reply: Thank you for this comment. Unfortunately, we are acquainted with binary logistic regression analysis but not acquainted with Poisson regression analysis. Also, since more than 12-hour delay is a sensitive issue for AMI patients in terms of clinical significance, we dichotomized the delay in ≤ 12 and > 12 hours to see what factors could contribute to > 12 hours delay. However, we have mentioned this issue in limitations

Comment: Women have different presentations of AMI, different social roles, and table 3 indicates that the median time to presentation is 21 hours. That’s astounding. Consider separating analysis into men and women, to see whether the associations differ.

Reply: We also thought about the different presentations of AMI between genders. We did not do a separate male and female analysis because of a relatively small sample size and also the numbers would be too small to obtain a meaningful difference between the genders. Although women with AMI had significantly longer delay in presentation to hospital in univariate analysis, they did not emerge as significant predictor in multivariate analysis. Additionally it will lengthen the result section of the study by a lot.

Comment: The authors describe this study as cross-sectional. However, the authors followed the patients by medical records in addition to the survey, so it would be reasonable to describe this as a retrospective cohort study. The investigators’ design is actually a very usual retrospective cohort study.

Reply: Thank you for this comment, another lesson learnt. We looked into the medical records for clinical presentation and comorbidities they had and did not follow the patients prospectively. We admit that the design is reasonably a retrospective cohort and changed it accordingly

Comment: The power is well-justified for a bivariate analysis. However, the authors do a multivariate regression analysis. The paper would be stronger with some type of sample size justification. Consider Gelman and Carlin’s Type S and Type M errors, at least to estimate probability that the estimates are in wrong direction.

Reply: We used the formula n=(z^2 p(1-p))/d^2 to justify the sample size. I am very sorry but we are junior researchers so we are not acquainted with Gelman and Carlin’s Type S and Type M errors.

Comment: Table 1 includes prior PCI and CABG, but this section seems like it should have also prior AMI as complementary information. — presumably this information was recorded and excluded as oversight?

Reply: This is a very valid point. We did plan to take previous Acute MI data. However, there were many patients from rural areas who aren’t well aware of their medical records and can’t provide medical records. It was hard to differentiate Acute MI or an angina episode from history alone when they did not undergo any procedure. However, they can definitely say if they underwent PCI and CABG, so we eventually decided to not use H/O Acute MI since the data would not be reliable.

Comment: The authors group smoking/tobacco into just one variable. Mode of ingesting tobacco may be important to report separately (e.g., bidi), and also betel/areca nut, if these were asked on survey.

Reply: Unfortunately, we generalized tobacco as a whole and did not separately inquire about smoking vs betel leaf consumption. I will keep this in mind for my future papers.

Comment: Table 2 gives just a few numbers and would be much more clear as a data display because clearly there’s a long tail on this distribution on time to arrive to hospital, and it’s very right-skewed. The information could be displayed in 4 histograms, for instance.

Reply: A very good point again, we have added 4 histograms accordingly also showing the skewness

Comments:

Tables should start on their own page and not run to the next.

Table 3 has a typo: under total column, percent of women is listed as 62.4%.

Reply: I apologize for these silly mistakes, these have now been corrected.

Comment: Age as dichotomized variable loses so much information. 3-5 groups may be more reasonable for age so the authors have a chance to detect an age effect.

Reply: We tried to do the analysis with 5 age groups but there was no still statistical significance so we just decided to show it as dichotomized to keep it consistent with the other variables in the table.

Comment: The p-value in table 3 is strictly to test the dichotomized version of the variable. However, a test of the quantitative variable would also be useful. It doesn’t need its own column: test the medians using Wilcoxon and use asterix or bold to indicate significance. I suspect 4 hour median difference by hypertension status is significant.

Reply: Unfortunately, to streamline the data collection process, we only collected data of the hypertension status of the patient but we do not have the quantitative data. Maybe we could focus on this and some other variables in a separate study in the future.

Comment: The authors find in table 3 that in general people with risk factors take longer to get to the hospital: people with hypertension, diabetes, family history, and smokers. Why is this? Is this association due to socioeconomic status? Age? The authors can evaluate this question with their data.

Reply: Thanks for this comment. We did not find any significant >12 hours delay with hypertension, family history and smokers. As for diabetes, we added another table to show why diabetes might have been causing this delay.

Comment: The regression has diagnosis as a predictor of delay, but diagnosis happened after the delay (after going to the hospital, that is) rather than before, so it’s not logical to put it into the regression. If there is a variable that is associated with the NSTEMI vs STEMI diagnosis that is apparent prior to going to the hospital, that would be more reasonable. In discussion authors mention less severe symptoms and say they can’t test. However, the symptom related variables collected in the survey may be associated with these

Reply: We did not include STEMI and NSTEMI in the regression.

“Diabetes Mellitus is not only a major risk factor of coronary artery disease, it also results in silent MI (25) and more atypical symptoms, all of which could contribute to delays in recognizing symptoms, and hence delays in presentation.” — this is interesting, and it’s a hypothesis that could be tested with the symptom data that was collected.

Reply: This was an excellent point. As I previously mentioned, we added an entire table showing the association of diabetes and type of presentation.

Submitted filename: Respone to reviewers.docx

Click here for additional data file.

13 Sep 2021

PONE-D-21-11112R1Pre-hospital delay and its associated factors in Acute Myocardial Infarction in a developing countryPLOS ONE

Dear Dr. Chowdhury,

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. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The reviewers commented favorably on your manuscript, but had some worthwhile suggestions. The authors should address the remaining issues. I am pleased to accept your manuscript, based on your revising it.

Please submit your revised manuscript by Oct 28 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're 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.

Please include the following items when submitting your revised manuscript:If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Michinari Nakamura, MD

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

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

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

**********

2. 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: Partly

**********

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

Reviewer #1: No

**********

4. 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

**********

5. 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

**********

6. 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: Authors made an excellent attempt at making all reviewer suggestions. Still a few minor things remain which affect overall interpretation and reading of the manuscript.

1) P values need to be removed from OR estimate sin the abstract and the results/discussion. You can leave them in the regression table, but it is most appropriate to interpret OR along with 95% CI's.

** You will notice for example that one of your variables was significant at the 0.03 level but the interval included 1, meaning although statistically significant it could not be as influential as others which did not include 1 in the interval.

2) Did you logistic regression include two outcomes? If not, then the regression that was conducted was a multivariable logistic regression, not a multivariate. Univariate logistic regression is a fine model selection to builds a final model, but realize that including all variables in a model may change the associations observed in univariate models. SOI in this way a forward or backward variable selection process maybe more appropriate than using univariate logistic regression to identify influential variables.

** The wording for multivariate needs changed. The other comment doesn't need included in the manuscript, other than to point out in the discussion that there are potential confounders not included in bivariate models.

3) Although the authors include a paragraph which mentions previous studies, they do not present specific findings from these in the introduction. This change is needed to enhance the readers experience and understanding for the current state of myocardial infarction in that country.

**********

7. 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

[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.]

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 PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

18 Oct 2021

Thanks to the reviewer for such a wonderful feedback. Once again, not only has these comments have helped this manuscript to be technically more sound, it also taught me a lot. We have attempted to revise as per the comments. I hope and pray that these revisions would suffice for the article to be eligible for publication.

Response to Reviewers:

Comment : 1) P values need to be removed from OR estimate sin the abstract and the results/discussion. You can leave them in the regression table, but it is most appropriate to interpret OR along with 95% CI's.

** You will notice for example that one of your variables was significant at the 0.03 level but the interval included 1, meaning although statistically significant it could not be as influential as others which did not include 1 in the interval.

Response: Thanks for this comment, another lesson learnt. The p-values have been removed accordingly from the text and only the ORs with their 95% CI have been mentioned.

Comment: 2) Did your logistic regression include two outcomes? If not, then the regression that was conducted was a multivariable logistic regression, not a multivariate. Univariate logistic regression is a fine model selection to builds a final model, but realize that including all variables in a model may change the associations observed in univariate models. SOI in this way a forward or backward variable selection process maybe more appropriate than using univariate logistic regression to identify influential variables.

** The wording for multivariate needs changed. The other comment doesn't need included in the manuscript, other than to point out in the discussion that there are potential confounders not included in bivariate models.

Response: I am sorry but as a junior researcher I am not sure whether I understood this comment correctly. Our logistic regression analysis included two outcomes so we kept the term “multivariate”. The wording for the multivariate regression analysis have been changed a little bit.

3) Although the authors include a paragraph which mentions previous studies, they do not present specific findings from these in the introduction. This change is needed to enhance the readers experience and understanding for the current state of myocardial infarction in that country.

Response: I apologize for not mentioning this. Findings have been added for reference 4 accordingly in introduction.

Submitted filename: Response to Reviewers2.docx

Click here for additional data file.

2 Nov 2021

Pre-hospital delay and its associated factors in Acute Myocardial Infarction in a developing country

PONE-D-21-11112R2

Dear Dr. Chowdhury,

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

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. 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 help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- 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.

Kind regards,

Michinari Nakamura, MD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

11 Nov 2021

PONE-D-21-11112R2

Pre hospital delay and its associated factors in Acute Myocardial Infarction in a developing country

Dear Dr. Chowdhury:

I'm 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 let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, 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.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Michinari Nakamura

Academic Editor

PLOS ONE