BACKGROUND

Patients discharged from hospitals confront significant vulnerability and excess rates of adverse events in the post-acute period.1, 2 Poor clinical outcomes during this dynamic time are attributed to many causes, including high-risk disease, evolving clinical needs, and fragmented, poor care coordination.3,4,5,6,7 Such factors are compounded in patients who lack access to fundamental medical and economic resources. Patients who are medically underserved, also referred to as low socioeconomic status (SES), broadly encompass a population subset that traditionally includes racial minorities, the chronically ill, homeless, and low-income populations who lack insurance coverage or have insufficient coverage through Medicaid.8 These patients defer necessary medical care, and present to the hospital sicker and with more advanced disease.9,10,11,12,13,14,15,16 Poor care access and cost have been identified within this population as major barriers to routine medical care, thereby increasing medical resource utilization.17,18,19,20,21,22

Despite an implicit understanding that the underserved population faces a multitude of unique struggles in the post-hospital period, the true nature of these barriers remains poorly described in the medical literature. In contrast, multiple studies in Medicare and commercially insured populations have examined the risk of hospital readmission and describe interventions to assist transitions in care to a primary care physician.23,24,25 As underserved patients comprise 30% of all patients, ineffectively addressing these issues in the peri-discharge period may have significant consequences on the overall quality and success of care delivery.26 Moreover, under the federal Hospital Readmissions Reduction Program (HRRP), safety-net and urban teaching hospitals caring for large percentages of patients of SES are disproportionately penalized for high readmission rates, reflecting financial consequences of this inadequate care.27 Research has shown broader health outcomes are strongly correlated to social determinants of health—environmental, social, and economic factors that are “conditions in which people are born, grow, live, work, or age.”28 These determinants directly account for over a third of total deaths in the USA each year. 29,30,31 As such, modification of these determinants may aid in achieving health equity between the underserved and insured populations.

In an effort to better identify unique barriers and to elucidate disparities in health outcomes among the medically underserved, we reviewed existing literature for specific characteristics of ineffective transitions following an episode of acute care. By categorizing these factors into an existing comprehensive framework, providers may be able to better influence and prioritize institutional policies in addressing these issues.28, 32

METHODS

We performed a literature review through Ovid MEDLINE and EMBASE covering January 1, 1960 to August 1, 2016, with the assistance of a research librarian. Ovid Medical Subject Headings (MeSH terms) were used to capture post-hospital care: “transition” and “patient discharge” or “patient readmission” with key word searches for “barrier,” “follow-up” (and all permutations), “rehospitalization,” “early discharge,” and “post-discharge.” To define the medically underserved population, the Medical Subject Headings “emigrants and immigrants,” “homeless persons,” “uninsured or underserved or underinsured,” “refugees,” “transients and migrants” and “Medicaid” were used, augmented with key word searches for low socioeconomic, uninsured, Medicaid, undocumented or homeless.

Inclusion Criteria

Articles were included in the review if they identified an underserved group primarily comprised of Medicaid, uninsured, or homeless adult persons and if the article specifically defined and described specific barriers to post-acute care. Only articles and studies performed in the USA were included. Articles and abstracts were excluded if they primarily addressed the commercial insurance population, Veterans, or Medicare recipients due to the increased services these groups have compared to the underserved. Studies focused on non-medical populations, e.g., surgical or psychiatric patients, were also excluded as to focus on establishment of a primary care medical home. Dual eligible Medicare-Medicaid were only included if the study specifically defined that population as underserved. There was no exclusion made based on study design, size, or methodological criteria. Only peer-reviewed articles were included. After the initial search, titles were reviewed to limit findings to adult medical patients based on the above inclusion criteria.

Data Synthesis

Selected full-text articles were read by the authors and determined for final inclusion by group consensus. After inclusion for final review, abstracted data elements included study design and type, number of patients, insurance status, article methodology, primary outcomes, and identified barriers. The listed barriers were then subcategorized by theme. Given the heterogeneous quality of the methods and outcomes among included studies, a meta-analysis or formal systematic review was not feasible.

RESULTS

The database search yielded 1702 articles. A total of 1635 records were excluded based on title or abstract, leaving 67 full-text articles for additional review (Fig. 1). Ultimately, 17 peer-reviewed papers met final inclusion criteria. The most common reasons for exclusion were lack of discussion of specific barriers to post-hospital follow-up and lack of payer source breakdown.

Fig. 1
figure 1

Flow diagram.

Full size image

The 17 included articles had heterogenous study designs32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48 (Table 1). Most were structured interviews (5) or direct patient surveys (5). Three articles were prospective cohorts 39, 41, 45 while three other articles were designed as a case-control study,47 a semi-structured interview,32 and a needs assessment34 respectively. Only one article was a randomized control study.38 The number of patients in each study ranged from 17 patients to 2974 (both structured interview methodology papers), with a median of 227 patients for all studies. Primary outcome metrics between studies varied greatly.

Table 1 Study Characteristics (N = 19)
Full size table

There was variability in how underserved populations were examined within studies. Six studies examined both Medicaid and uninsured populations32, 34, 36, 37, 40, 46, while one explored the uninsured only.47 One article studied Medicaid patients specifically,39 while another targeted the homeless population.42 The remaining eight articles defined the underserved population using methods other than payer source, labeled “all types.”33, 35, 38, 41, 43,44,45, 48

Cited determinants in obtaining post-hospital care included cost of medications or services, ability access to services, housing instability, transportation, health literacy, mental illness, and insufficient social support. Cost of care and lack of access were most frequently cited.

When regrouped within the construct of a framework of themes based on social determinants established in prior work, common themes included access failures and social fragility—including housing issues, lack of social support and basic resources, disease behavior, therapeutic misalignment, and accountability (Tables 2, 3 and 4).

Table 2 Study Descriptions
Full size table
Table 3 Barriers by Frequency of Discussion (Articles (N) = 17)
Full size table
Table 4 Barriers Categorized by Theme (Articles (N) = 17)
Full size table

DISCUSSION

Our literature review revealed a number of important findings relevant to hospital providers who care for medically underserved patients. We identified access to services and cost of health care as the most commonly cited factors influencing care transitions in this population. When reclassified into themes, access failures and social fragility were most commonly represented. This may reflect greater impediments to timely outpatient follow-up than previously realized in this population. Additionally, while patient accountability and substance abuse may impact an individual patient’s care, our findings highlight the critical role of tenuous social circumstances on discharged patients and their direct influence on health outcomes. Moreover, it seems unrealistic, if not impractical, to expect a medically underserved patient to effectively navigate his/her own health care needs given the high cost of care and often non-existent outpatient options available to them.

The results from this exhaustive review of 50 years of literature highlights the paucity of studies explicitly researching medically underserved populations and their unique obstacles to a successful post-hospital transition. Existing research evaluating the transitional period has been largely directed at specific diseases or Medicare populations.23, 24, 49, 50 Beyond the provision of suboptimal care, the omission of medically underserved patients with more unique care barriers from prior research may have serious cost implications given expanded Medicaid populations and hospital readmission provisions within the Affordable Care Act.51

The implications of our findings directly impact a number of important health care stakeholders and may inform and redirect strategies involving care coordination in this vulnerable population. Health care institutions should explore creative, targeted solutions that specifically improve access to essential care in the medically underserved population. For example, establishment of a medical home—frequently lacking in high-risk Medicaid patients—improves health, decreases care utilization and cost, and reduces disparities between socioeconomic subgroups.52, 53 Hospitals can improve access and reduce preventable acute care visits by coordinating timely access to publicly funded safety-net clinics.54,55,56,57 Another potentially beneficial strategy in connecting patients with a primary care provider is the initiation of home physician visits or telemedicine services.58 Programs allotting medications, transportation vouchers, and telephone reminders in the post-acute setting have also been shown to improve the likelihood of timely follow-up.59

To further enhance access to primary care and reduce hospital readmissions, hospitals should enlist the help of social and community supports, including community health workers.38, 60 On a larger scale, states or hospital systems may consider developing an integrated care coordination program that affords vulnerable populations a safety-net system, similar to those already active in Colorado, North Carolina, and Virginia. These programs provide comprehensive care management, contiguous patient care, and establishment of medical homes for vulnerable populations.61,62,63,64

Health care systems should also re-evaluate their discharge process when transitioning underserved patients to the outpatient setting. Historically, comprehensive discharge interventions broadly targeting all patients demonstrate low overall success, are frequently multicomponent and costly, and rarely address the unique or magnified factors within patients of low SES. 25, 65,66,67,68,69,70,71 Instead, providers tasked with discharging at-risk patients should focus on a patient’s ability to afford and execute the medical necessities in the context of fragile social support systems and housing, e.g., how likely is a patient being discharged able to afford his/her medications or obtain a ride to their scheduled appointment. Providers should engage the patient in shared decision-making regarding pharmaceutical costs prior to discharge, with realistic goal setting and transparency about differences in prices.72 Additionally, providers should establish the housing status of a discharging patient, as omitting this may impede the success of a patient with housing insecurity obtaining necessary medications, transportation, or mental health follow-up.42, 73 The use of checklists (e.g., SAFE DC) may provide a conceptual model when having these discussions in patients with identified housing insecurity.73

Targeted care management may additionally offer a solution for addressing social and economic barriers. One intervention on largely homeless inpatients at high risk for readmission demonstrated promise through integrated care management addressing housing, transportation, and social issues.74 Another hospital program utilized a patient needs assessment, serial stakeholder meetings, and formal programmatic development through integrating a nurse transitions advocate to assist patients in self-management and to link patients to outpatient care.62 Lastly, an intervention to provide respite care for homeless patients following hospitalization effectively reduced inpatient utilization.75

LIMITATIONS

Limitations of this paper include the significant heterogeneity of populations studied and an overall paucity of data represented. Such findings may represent ambiguity in defining this population, relative indifference from policy makers, and/or our narrow targeted interest in exploring studies specific to the post-acute care period. Additionally, the overall level of evidence of included studies was relatively low (given numerous retrospective and qualitative studies) with only one randomized controlled trial included. However, we posit inclusion of qualitative studies in this review is beneficial, if not necessary, as such studies represent the barriers as reported by patients themselves in describing their acute care transition experiences. In addition, absence of a reported barrier in an individual study did not preclude its existence, nor were individual barriers necessarily mutually exclusive, e.g., a patient who reported difficulty with affording medications may not have separately reported difficulty obtaining transportation. Likewise, other barriers (e.g., educational level, racism) may exist beyond those explored in our included studies. Lastly, we were not able to quantify the relative impact of the various defined barriers on care transitions due to the nature of the included studies.

CONCLUSIONS

Expansion of Medicaid and broad healthcare reform is not sufficient to ensure improved population health outcomes without concurrent interventions that address socioeconomic issues that consistently plague vulnerable patients. While multiple, large-scale health policy initiatives have addressed care coordination among Medicare beneficiaries, unique transitional care barriers and effective interventions among Medicaid and uninsured populations remain largely unaddressed.76 Additional research into the real-world issues that negatively affect at-risk patients is needed to gain broader insight and to create effective tools to improve the health of this vulnerable community and reduce health care costs.