Patient Perspectives on Care Transitions From Hospital to Home.

Importance: Understanding the patient's perspective of their care transition process from hospital or skilled nursing facility (SNF) to home may highlight gaps in care and inform system improvements. Objective: To gather data about patients' care transition experiences and factors associated with follow-up appointment completion. Design, Setting, and Participants: A survey tool was developed with input from patient advisors and organizations participating in a collaborative quality initiative. Seventeen hospitals, 12 practitioner organizations, and 6 SNFs in Michigan collaborated to identify shared patients who were aged 18 years and older, had a working telephone number, recently returned home or to an assisted living facility with a diagnosis of congestive heart failure or chronic obstructive pulmonary disease, or after an SNF stay. Using consecutive sampling, interviewers collected 5 telephone surveys per month. From October 2018 to December 2019, patients or caregivers were surveyed via telephone 8 to 12 days after discharge from a hospital or SNF. Data were analyzed from March 2020 to January 2022. Exposure: Care transition experiences. Main Outcomes and Measures: The primary outcome was to identify patient-perceived gaps during care transition experiences, including postdischarge follow-up.
Results: On the basis of pilot data, the response rate was estimated at 34%, yielding 1257 surveys. Of 1257 survey respondents (mean [SD] age, 70 [12.94] years for 968 patients for whom age data was available), 654 (52%) were female; 829 (74%) were White, 250 (22%) were Black or African American, and 40 (4%) were another race. Eleven percent of patients reported not receiving a telephone number to call for postdischarge questions. Nearly 80% of patients (977 patients) received a follow-up telephone call, and most found it valuable. Twenty percent of patients (255 patients) reported at least 1 social determinant of health issue. Lack of transportation was associated with reduced likelihood of completing a follow-up visit, decreasing the odds of completing a follow-up by nearly 70% (odds ratio [OR], 0.31; 95% CI, 0.18-0.53; P < .001). Compared with other patient groups, Black patients were less likely to report completing a postdischarge follow-up visit (OR, 0.49; 95% CI, 0.36-0.67; P < .001) or to receive prescribed medical equipment (OR, 4.23; 95% CI, 1.30-13.83; P = .02). Conclusions and Relevance: An examination of patient discharge experiences from a hospital or SNF identified inconsistencies in care transition processes, social determinants of health issues needing to be addressed after discharge, and racial disparities between patients who attend follow-up appointments. Physicians should be aware of these findings and their consequences for patient experiences.

Introduction

Health systems aim to improve the quality and value of care by improving patient experiences and clinical outcomes while minimizing costs. One potential route to improving value is improving the transition of care process, which begins before the patient’s discharge and continues beyond their return home. The process requires many elements, including education, medication reconciliation, follow-up appointments and telephone calls, as well as supportive care once home.[1,2,3] Identifying and addressing patients’ needs after hospitalization can improve health and minimize hospital reutilization.[4,5] A patient’s health can be affected by confusion, gaps in care, lack of resources, and poor information transfer between physicians and patients at hospital discharge.[6,7,8] Efforts are focused on the care transition process because of a 1 in 5 chance of adverse events, such as a readmission or emergency department visit, within 30 days following hospital discharge.[9] Past studies[10] have evaluated parts of the care transition process and the patient’s perspective. For example, the Centers for Medicare & Medicaid Services Hospital Consumer Assessment of Healthcare Providers and Systems features care transition questions but has a narrow focus, leaving facets of the care transition experience unaddressed. Collecting patient-reported outcomes (PROs) is another way to gather more information regarding a patient’s care transition experience and develop a deeper understanding of many issues patients cope with when they leave the hospital or skilled nursing facility (SNF), including social determinants of health (SDOH) and other disparities. Although patient perceptions can be skewed by individual expectations,[11] their experiences can still inform care transition improvements. In this study, we share findings regarding care transition processes, SDOH concerns, postdischarge follow-up, and racial disparities across patient experiences.

Methods

This survey study followed the American Association for Public Opinion Research (AAPOR) reporting guideline. The Integrated Michigan Patient-Centered Alliance in Care Transitions (I-MPACT) works with hospital and practitioner organization partnerships (or clusters) to improve care transitions for patients discharged from the hospital to home with congestive heart failure (CHF) or chronic obstructive pulmonary disease (COPD) and patients with any condition discharged from the hospital to a skilled nursing facility (SNF). A PROs survey was developed to capture details about patients’ experiences within 8 to 12 days after discharge from a hospital or SNF and was conducted by telephone interviewers from participating I-MPACT clusters (eg, nurses, clerks, and care managers). The intent of the survey was to inform ongoing quality efforts to increase the volume and completion of 7-day postdischarge follow-up visits and improve patient care transitions. The institutional review board at the University of Michigan reviewed I-MPACT’s quality improvement efforts and provided a determination of exempt, including exemption from informed consent for our survey because no additional patient identifiers were included and consent is implied when the patient/caregiver agrees to do the survey.

Survey Development

With no validated survey instrument available, a structured questionnaire was developed by I-MPACT with input from local patient advisors and frontline representatives from clusters. Many factors were considered in survey development, including adherence to Centers for Medicare & Medicaid Services requirements to avoid overlap with Hospital Consumer Assessment of Healthcare Providers and Systems questions, using plain language for patient understanding, creating uniformity across survey questions, and using primarily closed-ended questions. Because the patient population was older and many had chronic diseases, adult caregivers were allowed to answer survey questions on the patient’s behalf. Respondents were asked Likert-scale and open-ended questions about patients’ postdischarge experiences, including attendance at postdischarge follow-up appointments, number of follow-up telephone calls received, discussions with physicians about personal health care goals, concerns with medical equipment, preparation for discharge, and SDOH concerns (see the eAppendix in the Supplement). SDOH issues assessed included patient concerns with lack of transportation, affording prescriptions, use of medical equipment, affording doctor visits and copayments, meeting basic needs, and getting help at home. If respondents did not recall an answer or preferred not to answer, a question could be skipped. Patients self-reported age and race. Race categories were White, Black or African American, American Indian or Alaska Native, Asian, Native Hawaiian or other Pacific Islander, or any other race. Race was assessed in this study to capture any disparities in patient experience that might be associated with race. After the telephone interview ended, interviewers reviewed the patient’s electronic health record and answered mandatory questions such as who (patient or caregiver) answered the survey questions, patient target population, insurance, and sex. Interviews were considered complete after respondents were asked all patient interview questions and mandatory questions were answered.

The survey script was piloted for 3 months with minimal modifications made to the survey instrument. The pilot response rate was 33.9% (369 calls resulting in 125 interviews). Surveys were collected using consecutive sampling from October 2018 through December 2019. Each cluster completed 5 interviews monthly.

Clusters were required to keep a call log including both complete and incomplete surveys. Using the logs, audits were conducted to confirm respondents were aged 18 years or older and surveys occurred between 8 and 12 days after discharge. Logs were not retained because of protected health information restrictions, which prevented the calculation of the survey response rate. Of the 1257 surveys collected, 968 were completed by patients and 289 by their caregivers. Patient age was not collected for the 289 surveys completed by a caregiver but was collected for the remaining 968 patients. Race was not collected for 19 patients whose surveys were completed by a caregiver and for 119 patients who completed the surveys by themselves.

Statistical Analysis

We described the patient sample using demographic information including sex, race, age, and insurance status. We also reported the proportion of patients who experienced SDOH issues, received postdischarge follow-up calls, and attended follow-up visits. We reported these statistics for the overall sample and for groups stratified by Black and White or other race (other race included American Indian or Alaska Native, Asian, Native Hawaiian or other Pacific Islander, or some other race) and used χ2 tests to evaluate group differences (Table 1). We estimated multivariate logistic models for SDOH and transition of care factors that had a significant racial gap in Table 1, with models adjusted for demographic variables. We also assessed patient demographics, SDOH, and care transition factors that might be associated with completion of postdischarge follow-up visits (Table 2). Finally, we estimated a multivariate logistic regression model to further evaluate racial disparity gaps in postdischarge follow-up completion observed in bivariate analysis. All demographic and other factors having a significant, bivariate association (P < .10) with completion of postdischarge follow-up were used as covariates in the multivariate model. As some responses on race (138 respondents [11%]) and age (289 respondents [23%]) were missing for the sample, each multivariate model was estimated using full information maximum likelihood (FIML). FIML avoids listwise deletion of observations and uses all available data points in model estimation.[12] For comparisons, each multivariate model was also estimated using complete case analysis. The article only reports estimates from models using FIML (Table 3), as these estimates are more robust than estimates from models using listwise deletion.[13] The statistical significance was determined at 2-sided P < .05, and all analyses were performed using Stata statistical software release 16[14] (StataCorp) and Mplus software version 8 (Muthén & Muthén).[15] Data were analyzed from March 2020 to January 2022.

Table 1.

Patient Demographics, SDOH, and Transition of Care Factors Overall and by Race

Demographics	Patients, No. (%)			P value
	Overall (N = 1257)	White or other (n = 869)	Black or African American (n = 250)
Race
Black or African American	250 (22)	0	250 (100)	NA
White	829 (74)	829 (95.4)	0
Other raceb	40 (4)	40 (4.6)	0
Total	1119	869 (100)	250 (100)
Sex
Female	654 (52)	455 (52)	114 (46)	.06
Male	603 (48)	414 (48)	136 (54)
Total	1257	869	250
Age, y
21-64	312 (32)	175 (28)	86 (40)	.001
65-74	306 (31)	204 (32)	70 (32)
75-84	218 (23)	152 (24)	45 (21)
≥85	132 (14)	101 (16)	16 (7)
Total	968	632	217
Insurance type
Medicare	901 (72)	650 (75)	169 (68)	.07
Medicaid	71 (6)	48 (5)	20 (8)
Other (commercial, preferred provider organization, or health maintenance organization)	285 (22)	171 (20)	61 (24)
Total	1257	869	250
Living situation
Lives at home				<.001
With others	836 (67)	604 (70)	154 (62)
Alone	367 (30)	225 (26)	93 (37)
Other	42 (3.4)	36 (4)	3 (1)
Total	1245	865	250
Target population
Congestive heart failure	890 (71)	549 (63)	210 (84)	<.001
Chronic obstructive pulmonary disease	191 (15)	160 (18)	27 (11)
Skilled nursing facility	133 (11)	121 (14)	12 (5)
Other	43 (3)	39 (5)	1 (<1)
Total	1257	869	250
SDOH
Cannot afford medications
Yes	93 (8)	55 (6)	24 (10)	.07
No	1134 (92)	802 (94)	220 (90)
Total	1227	857	244
Cannot afford medical visits
Yes	69 (6)	37 (4)	23 (9)	.002
No	1158 (94)	820 (96)	221 (91)
Total	1227	857	244
Cannot afford basic needs
Yes	36 (3)	18 (2)	12 (5)	.02
No	1191 (97)	839 (98)	232 (95)
Total	1227	857	244
Lack of transportation
Yes	74 (6)	46 (5)	18 (7)	.24
No	1153 (94)	811 (95)	226 (93)
Total	1227	857	244
Not enough help at home
Yes	66 (5)	43 (5)	12 (5)	.95
No	1161 (95)	814 (95)	232 (95)
Total	1227	857	244
Other SDOH
Yes	27 (2)	21 (3)	4 (2)	.45
No	1200 (98)	836 (97)	240 (98)
Total	1227	857	244
No. of SDOH
≥1	255 (21)	160 (19)	57 (23)	.10
0	972 (79)	697 (81)	187 (77)
Total	1227	857	244
Transitions of care
Given name or telephone number to call
Yes	1097 (89)	757 (89)	203 (83)	.01
No	141 (11)	97 (11)	43 (17)
Total	1238	854	246
Was not able to take medications as prescribed
Yes	72 (6)	50 (6)	14 (6)	.92
No	1174 (94)	810 (94)	234 (94)
Total	1246	860	248
Prepared to go home
Yes	1063 (86)	744 (86)	217 (87)	.31
No, but it did not affect health	121 (10)	77 (9)	24 (10)
No, but it did affect health	54 (4)	44 (5)	7 (3)
Total	1238	865	248
Confident using medical equipment
Yes	814 (92)	578 (94.4)	148 (92.5)	.04
No	68 (8)	34 (5.6)	12 (7.5)
Total	882	612	160
Did not receive equipment
Yes	17 (2)	8 (1)	7 (4)	.02
No	882 (98)	612 (99)	160 (96)
Total	899	620	167
Calls received after discharge
None	266 (21)	202 (24)	52 (21)	.83
1-2	589 (47)	416 (48)	125 (50)
3-4	277 (22)	175 (20)	54 (21)
≥5	111 (9)	67 (8)	19 (8)
Total	1243	860	250
Calls helpful
Yes	867 (90)	589 (91)	171 (87)	.07
No	97 (10)	57 (9)	26 (13)
Total	964	646	197
Talked to physician about personal health goals
Yes	894 (73)	598 (70)	180 (75)	.15
No	335 (27)	258 (30)	61 (25)
Total	1229	856	241
Postdischarge follow-up
Follow-up				<.001
Follow-up completed	772 (63)	568 (67)	125 (52)
Scheduled follow-up but not yet completed	343 (28)	220 (26)	84 (35)
Not scheduled or completed	105 (9)	60 (7)	32 (13)
Total	1220	848	241
Completed postdischarge follow-up
Yes	772 (63)	568 (67)	125 (52)	<.001
No	448 (37)	280 (33)	116 (48)
Total	1220	848	241

Abbreviations: NA, not applicable; SDOH, social determinants of health.

P values are based on χ2 tests comparing the distribution of patient factors in the 2 racial categories.

Other race included Asian, American Indian or Alaskan Native, Native Hawaiian or other Pacific Islander, and any other race.

Table 2.

Factors Associated With Completion of Postdischarge Follow-up (Unadjusted)

Factors	Patients, No. (% who completed follow-up) (N = 1220)	P valuea
Demographics
Race
Black or African American	241 (52)	<.001
White or other	848 (67)
Total	1089
Sex
Female	629 (64)	.55
Male	591 (62)
Total	1220 (860)
Age, y
21-64	309 (56)	.004
65-74	296 (65)
75-84	213 (71)
≥85	127 (63)
Total	945
Lives at home with others
Yes	820 (67)	<.001
No	394 (56)
Total	1214
Insurance type
Medicare	877 (66)	.009
Medicaid	71 (51)
Other (commercial, preferred provider organization, or health maintenance organization)	272 (59)
Total	1220
Target population
Congestive heart failure	861 (66)	<.001
Chronic obstructive pulmonary disease	187 (50)
Skilled nursing facility	132 (60)
Other	40 (70)
Total	1220
SDOH
Cannot afford medications
Yes	91 (67)	.21
No	1104 (52)
Total	1195
Cannot afford medical visits
Yes	68 (59)	.44
No	1127 (63)
Total	1195
Cannot afford basic needs
Yes	35 (46)	.03
No	1160 (64)
Total	1195
Lack of transportation
Yes	73 (33)	<.001
No	1122 (65)
Total	1195
Not enough help at home
Yes	62 (52)	.05
No	1133 (64)
Total	1195
Other SDOH
Yes	26 (58)	.56
No	1169 (63)
Total	1195
≥1 SDOH
Yes	249 (51)	<.001
No	946 (66)
Total	1195
Transition of care
Given name or telephone number to call
Yes	1066 (64)	.19
No	135 (59)
Total	1201
Calls received after discharge
Yes	953 (64)	.05
No	253 (58)
Total	1206

Abbreviation: SDOH, social determinants of health.

P values are based on χ2 tests.

Table 3.

Association of Demographic Characteristics With SDOH, Transition of Care, and Completion of Postdischarge Follow-up

Outcome and category	Multivariate logistic model using FIMLa
	OR (95% CI)	P value
Cannot afford basic needs, Black or African American	1.5 (0.7-3.4)	.29
Cannot afford physician visits, Black or African American	1.6 (0.9-2.9)	.12
Given name or telephone number to call, Black or African American	0.5 (0.3-0.7)	.001
Did not receive equipment Black or African American	4.2 (1.3-13.8)	.02
Completion of follow-up visit
Black or African American	0.5 (0.4-0.6)	<.001
Female	1.1 (0.9-1.5)	.33
Age, y
21-64	1.3 (0.8-2.3)	.26
65-74	1.5 (0.9-2.4)	.09
75-84	1.9 (1.1-3.1)	.02
Insurance type
Medicare	1.3 (0.8-2.3)	.33
Other (commercial, preferred provider organization, or health maintenance organization)	1.1 (0.6-1.9)	.74
Lives at home with others	1.5 (1.2-1.9)	.002
Target population
Chronic obstructive pulmonary disease	.04 (0.3-0.6)	<.001
Skilled nursing facility	0.7 (0.5-1.1)	.10
Other	0.9 (0.4-1.8)	.69
SDOH
Cannot afford basic needs	0.8 (0.4-1.5)	.44
Lack of transportation	0.3 (0.2-0.5)	<.001
Not enough help at home	0.7 (0.4-1.2)	.16
Transitions of care
Calls received after discharge	1.3 (1.0-1.8)	.06

Abbreviations: FIML, full information maximum likelihood; OR, odds ratio; SDOH, social determinants of health.

Reference categories used in multivariate logistic model were age 85 years and older, enrollment in Medicaid, and congestive heart failure diagnosis. Each of the 4 multivariate logistic models was adjusted for sex, age, insurance type, living situation, and target population. Model estimates for covariates are not shown.

Results

Patient Demographics

Of 1257 total patients in the sample, 654 (52%) were female, 829 (74%) were White, 250 (22%) were Black, and 40 (4%) belonged to another race. Patient age ranged from 21 to 99 years with a mean (SD) of 70 (12.94) years for the 968 patients for whom age data was available. Differences in demographic characteristics were observed among White and Black patients: Black patients were more likely to be younger (χ23 = 16.94; P < .001) and less likely to live at home with others (χ22 = 15.31; P < .001). The primary diagnosis categories for patients were CHF (890 patients [71%]), COPD (191 patients [15%]), patients discharged from hospital to a SNF with any diagnoses (133 patients [11%]), and other conditions (43 patients [3%]).

Patients’ Transitions of Care Experiences

Responses point to several gaps in the care transition process. More than 1 in 10 patients (141 patients [11.4%]) reported not being provided a telephone number to call regarding questions about their care following hospital or SNF discharge (Table 1). Over 20% of patients reported not receiving a follow-up telephone call. A higher proportion of Black patients indicated they did not receive a telephone number to call with questions compared with White patients and patients of other races (17.5% [43 patients] vs 11.4% [97 patients]; χ22 = 15.31; P = .01). This racial gap remained after adjusting for demographic variables (odds ratio [OR], 0.49; 95% CI, 0.32-0.75; P = .001). Additionally, among patients who required medical equipment at home (eg, oxygen, inhalers, or blood pressure monitors), 7.7% (68 patients) reported lack of confidence using such equipment, and 1.9% (17 patients) reported never receiving prescribed equipment. A higher proportion of Black patients than White patients and patients of other races reported not receiving prescribed equipment (4.2% [7 patients] vs 1.3% [8 patients]; P = .02), and this association remained significant in multivariate analysis (OR, 4.23; 95% CI, 1.30-13.83; P = .02). Patient-reported experiences also point to care gaps in the critical period following a hospital or SNF discharge. Although a large proportion of patients (977 [78.6%]) reported receiving at least 1 follow-up call within 8 to 12 days of hospital or SNF discharge, and 8.9% (111 patients) reported 5 or more calls, a sizable proportion (266 patients [21.4%]) reported not receiving any follow-up calls. Although a slightly higher proportion of Black patients reported receiving a follow-up call than White patients and patients of other races (198 patients [79.2%] vs 658 patients [76.5%]), a higher proportion of Black patients than White patients and patients of other races reported that the call(s) were not helpful (26 patients [13.2%] vs 57 patients [8.8%]). However, these associations were not significant. Overall, 867 patients (89.9%) found calls helpful or very helpful.

SDOH Characteristics

Notably, 1 in 5 patients (255 patients [20.8%]) reported at least 1 SDOH concern. The 4 most common concerns were (1) affording components of care such as prescriptions, medical equipment, physical therapy, and home health care (93 patients [7.6%]); (2) access to transportation for phyisician appointments, pharmacy, grocery store, and so forth (74 patients [6.0%]); (3) affording medical visits and copayments (69 patients [5.6%]); and (4) having enough help at home to care for oneself (66 patients [5.4%]). A higher proportion of Black patients than White patients and patients of other races reported concerns affording doctor visits and copayments (χ21 = 9.62; P = .002) and affording basic needs such as food and heat (χ21 = 5.69; P = .02) (Table 1). However, in adjusted models (Table 3), a significant racial gap was not observed for concerns about affording physician visits or affording basic needs.

Completion of Postdischarge Follow-up

When surveyed, 63.3% of patients (772 patients) reported having seen a physician, and another 28.1% (343 patients) reported having an appointment scheduled. However, data showed substantial race and ethnicity gaps in postdischarge follow-up, with Black patients more likely than White patients and patients of other races to not have an appointment scheduled (32 patients [13.3%] vs 60 patients [7.1%]), and less likely to have seen a physician at the time of the survey (125 patients [51.9%] vs 568 patients [67.0%]) (Figure).

Figure.

Postdischarge Follow-up Completion Rates for Patients Overall and by Race

The χ2 test was used to compare differences in postdischarge follow-up completion among 1220 White or other race and Black patients (P < .001). Other race included Asian, American Indian or Alaskan Native, Native Hawaiian or other Pacific Islander, and any other race.

To determine factors associated with completion of a physician visit, we compared patients who reportedly attended a physician visit to those who did not (Table 2). Patient race, age, insurance status, admission diagnosis, and living situation were associated with completion of physician follow-up. Among SDOH, inability to afford basic needs and lack of transportation were barriers to completion of physician follow-up. Additionally, patients who reported at least 1 SDOH concern were less likely to complete a physician visit than patients who reported no SDOH concerns (Table 2).

We estimated a multivariate model to further assess the racial gap in completion of physician visits. According to the model (Table 3), Black patients had a lower likelihood of follow-up completion than White patients and patients of other races (OR, 0.49; 95% CI, 0.36-0.67; P < .001). Patients who reported transportation concerns had a lower likelihood of follow-up completion than patients who did not report such concerns (OR, 0.31; 95% Cl, 0.18-0.53; P < .001). Patients living at home with others had a higher likelihood of follow-up completion (OR, 1.52; 95% Cl, 1.17-1.97; P = .002) compared with patients reporting other living situations. As seen in Table 3, patients with COPD (OR, 0.45; 95% Cl, 0.32-0.64; P < .001) had a lower likelihood of follow-up completion than patients with CHF. Patients who received 1 or more follow-up calls had a higher OR for completion of a physician visit than patients who did not (OR, 1.33; 95% CI, 0.99-1.79; P = .06), although the difference was not significant. Patients living with others had 50% increased odds of completing follow-up, whereas those with a lack of transportation had a 70% lower odds of completing a visit.

Discussion

Prior PROs studies[4,16] have collected information on sufficient support at home, recall of instructions for self-care, a number to call with concerns, ability to manage medications correctly, and postdischarge follow-up appointments, but none that we know of have covered multiple aspects of care transition processes, SDOH concerns, and follow-up. In this study, I-MPACT surveyed patients about their perceptions of care transition processes and any SDOH affecting them during the postdischarge period, to determine any associations among demographics, care transition processes, or SDOH, and completion of 7-day follow-up.

Overall, these findings show that most patients receive postdischarge follow-up telephone calls and find them valuable, but 21% of patients do not receive a telephone call, indicating inconsistencies in care transition processes. Although clinicians anecdotally feel that patients are bombarded by too many calls, PROs responses show 89.9% of patients found calls helpful or very helpful. For example, in open-ended question responses, patients provided feedback such as, “Nurse calling to explain my diet (CHF) in more detail and how to limit fluid intake was helpful,” and “Pharmacy called me to go over my medications again after I returned home. That was helpful.” Other studies have shown that patients’ recollection of discharge instructions is incomplete.[3,17] Our study emphasizes the need for postdischarge contact with patients to address any gaps of care. Although most patients find follow-up calls helpful, there was a disparity in the number of Black patients finding follow-up calls unhelpful vs White patients and patients of other races. We are unable to assess whether this disparity is associated with varied patient expectations, racial bias, or other factors.

We also found that 1 in 10 patients reported not receiving a telephone number to call regarding their care after discharge, with a higher proportion of Black patients not receiving a telephone number to call. In addition, Black patients reported not receiving prescribed equipment more often than White patients, and these gaps persisted even after adjustment for demographic variables. These care transition processes continue to be flawed from patients’ perspectives and can lead to repeat hospital utilization.

One in 5 patients surveyed for PROs reported SDOH concerns, such as the inability to afford prescriptions, medical care, doctor appointments, and basic needs; transportation issues; and having adequate assistance at home. Although the health care industry is aware of the important role SDOH plays in patient health,[18] awareness has not translated into improvement. In a 2019 survey of Michigan seniors, 34.7% noted their reason for not seeing a physician for follow-up was because they could not afford to, another 18.1% did not because of lack of insurance coverage, and 22.1% did not because of lack of transportation.[19] It is notable that PROs interviewers were reluctant to ask SDOH questions because of the possibility that social issues would be identified they did not feel equipped to address.

In terms of postdischarge follow-up, most patients reported having seen a physician, with nearly another one-third reporting they had an appointment scheduled. Compared with other studies of postdischarge follow-up, our rates of follow-up are high.[4,7,20] However, we found Black patients reported fewer scheduled or completed follow-ups with physicians compared with White patients and patients of other races. Multivariate analysis also found Black patients had a lower likelihood of follow-up completion compared with White patients and patients of other races, even after adjusting for other covariates. Patients living with others had 50% increased odds of completing follow-up, whereas those with a lack of transportation had a 70% lower odds of completing a visit. Increased availability of virtual visits may help patients with completion of early follow-up visits.

Limitations

This study has limitations, including possible selection bias. Only patients or caregivers who were at home or assisted living facilities with a telephone number on record were included. Those who answered surveys may have different perceptions of the care transition process or report fewer SDOH issues, leading to an underestimate of SDOH issues during the postdischarge period. AAPOR guidelines were followed for many aspects of the survey development including planning survey questions, training of interviewers, piloting (pretesting) the questionnaire and procedures, constructing quality checks, and using statistical analytic and reporting techniques, but AAPOR guidelines for sampling and reporting for survey studies were not used, as monthly consecutive sampling occurred at each location according to discharge date. In addition, although response rate was calculated for the pilot (33.9%), it was not captured during the main survey. Because of low incidence, our finding associated with racial disparities in receiving medical equipment may have low stability and should be replicated in future studies with a larger sample. Future PROs tools should also evaluate additional reasons for noncompletion of physician visits, such as time away from work, lack of an assigned primary care physician, doubts about benefits of a primary care visit, and mistrust of health care systems and/or physicians.[21,22]

We were unable to determine whether the presence of SDOH concerns, problems with medication or equipment, follow-up telephone calls, race, or any other factors were associated with increased readmissions. However, there is growing evidence from the Medicare-funded Community-based Care Transition Programs that interventions addressing social needs can decrease unnecessary hospital utilization.[23]

Our analysis showed a substantial racial disparity in completion of follow-up appointments. Unfortunately, our study was not designed to capture the reasons why Black patients attend fewer follow-up appointments after discharge. In addition to disparities in follow-up appointments attended, Black patients were less likely to receive the number of someone to call with questions after discharge and were less likely to receive their medical equipment. Some of this disparity could be due to health literacy and access factors; however, we could not account for these in our study. In this study and others, the existence of racial and ethnic disparities persists even when adjusting for income, age, and insurance status.[24] Although research shows that disparate levels of insurance and health care access are associated with racial disparities in health care, these disparities have decreased since the implementation of the Patient Protection and Affordable Care Act.[25,26] More health systems need to focus quality improvement work on addressing racial disparities to improve patient care, reduce social and structural inequities in health delivery, and focus on population health.[27,28]

Conclusions

Results from PROs shed light on vulnerabilities for patients when leaving the hospital or SNF. There are still multiple opportunities for improvement, including (1) providing reliable, systematic care transition processes for all (follow-up telephone calls, numbers for patients to call, and delivered home medical equipment); (2) addressing patient SDOH, such as transportation; (3) scheduling and helping patients attend follow-up appointments; and (4) recognizing and reducing racial disparities in care. This information on patient challenges during the transition of care process could help hospitals and physicians tailor future care transition interventions to be specific to their patients’ needs. The data provided by the I-MPACT PROs survey are a rich source of actionable information that can inform changes in patient care transition processes, reduce racial disparities, and improve patient care and outcomes.