Currently Viewing:
The American Journal of Managed Care September 2017
Guideline Concordance of New Statin Prescriptions: Who Got a Statin?
Thomas Cascino, MD; Marzieh Vali, MS, BS; Rita Redberg, MD, MSc; Dawn M. Bravata, MD; John Boscardin, PhD; Elnaz Eilkhani, MPH; and Salomeh Keyhani, MD, MPH
Provider-Owned Insurers
David H. Howard, PhD, and Erin Trish, PhD
The Effect of Narrow Network Plans on Out-of-Pocket Cost
Emily Meredith Gillen, PhD; Kristen Hassmiller Lich, PhD; Laurel Clayton Trantham, PhD; Morris Weinberger, PhD; Pam Silberman, JD, DrPh; and Mark Holmes, PhD
In-Gap Discounts in Medicare Part D and Specialty Drug Use
Jeah Jung, PhD; Wendy Yi Xu, PhD; and Chelim Cheong, PhD
Racial and Ethnic Differences in Hip Fracture Outcomes in Men
Lucy H. Liu, MD, MPH; Malini Chandra, MS, MBA; Joel R. Gonzalez, MPH, MPP; and Joan C. Lo, MD
Integrating Behavioral Health Under an ACO Global Budget: Barriers and Progress in Oregon
Jason Kroening-Roché, MD, MPH; Jennifer D. Hall, MPH; David C. Cameron, BA; Ruth Rowland, MA; and Deborah J. Cohen, PhD
Evaluation of a Packaging Approach to Improve Cholesterol Medication Adherence
Hayden B. Bosworth, PhD; Jamie N. Brown, PharmD, BCPS; Susanne Danus, BS; Linda L. Sanders, MPH; Felicia McCant, MSSW; Leah L. Zullig, PhD; and Maren K. Olsen, PhD
Treatment Barriers Among Younger and Older Socioeconomically Disadvantaged Smokers
Patrick J. Hammett, MA; Steven S. Fu, MD, MSCE; Diana J. Burgess, PhD; David Nelson, PhD; Barbara Clothier, MS, MA; Jessie E. Saul, PhD; John A. Nyman, PhD; Rachel Widome, PhD, MHS; and Anne M. Joseph
Currently Reading
Against the Current: Back-Transfer as a Mechanism for Rural Regionalization
Leah F. Nelson, MD, MS; Karisa K. Harland, PhD, MPH; Dan M. Shane, PhD; Azeemuddin Ahmed, MD, MBA; and Nicholas M. Mohr, MD, MS

Against the Current: Back-Transfer as a Mechanism for Rural Regionalization

Leah F. Nelson, MD, MS; Karisa K. Harland, PhD, MPH; Dan M. Shane, PhD; Azeemuddin Ahmed, MD, MBA; and Nicholas M. Mohr, MD, MS
The authors investigated back-transfer: the transfer of patients near the end of an acute hospitalization to a local community hospital for completion of their medical care.
ABSTRACT

Objectives: This paper investigates back-transfer: the transfer of patients near the end of their acute hospitalization to a local community hospital for the completion of their medical care. We seek to describe factors contributing to back-transfer, with the goal of elucidating the current use of back-transfer and barriers to its more widespread adoption for rural healthcare regionalization.

Study Design: Observational unmatched case-control. 
 
Methods: This was a retrospective study of adults hospitalized in Iowa between 2005 and 2013 to identify back-transferred patients. Demographic, geographic, rurality, procedural, and disease information was compared among cases and control groups using univariate analysis and multivariable logistic regression. 

Results: Over the 9-year period, 172,544 back-transfer eligible patients were admitted to 1 of 5 large Iowa hospitals, of which 287 (0.2%) were back-transferred. Back-transferred patients were more likely than their non–back-transferred counterparts to be older, male, and white; to live in large rural areas; and to have public insurance. As inpatients, they had longer median lengths of stay (15 vs 5 days; P <.001), more medical comorbidities, and were more likely to have a cardiac catheterization procedure than the control group.

Conclusions: Back-transfer is a very rare event. While demographic and medical differences between back-transferred patients and controls may partially explain the infrequency, other systematic barriers must exist to limit back-transfer. These barriers likely include legal, financial, logistical, and patient care concerns. Despite the rarity with which it is employed, back-transfer is a promising strategy that could better utilize health resources, especially in rural America. 

Am J Manag Care. 2017;23(9):e287-e294
Takeaway Points

Our study found that back-transfer of patients from large hospitals to small hospitals during an acute care episode is a rare event. Demographic and medical differences may partially explain the infrequency, but systematic (legal, financial, logistical, and patient care) barriers must exist to limit back-transfer. 
  • Back-transport could maximize availability of specialty services, support rural health systems, and move patients closer to home and family. 
  • Widespread adoption of back-transfer through changes in policy and clinical practice patterns may be a way to better utilize healthcare resources.
The concepts of regionalization and centralization are central to managed care. Here, we define regionalization as an “active process by which patients are appropriately matched to appropriate resources.”1 This can be compared with centralization, which is a more unplanned process in which patients are transferred to larger medical centers for a variety of medical, financial, legal, and personal reasons. In rural states, regionalization and centralization processes occur concurrently. Many small hospitals are closing and, thus, care is becoming centralized to fewer hospitals. Simultaneously, small local hospitals are unable to provide comprehensive services, so patients are often transferred for specific types of problems, such as stroke care and cardiac catheterization.

Although transfer to regional care centers is lifesaving for many rural residents, variability exists in patient preferences for transfer. Liu et al reported that 56% of rural patients who lived near a federally designated critical access hospital (CAH) would have preferred to have care locally, but were transferred or referred by their local provider because the care they required was unavailable at their local CAH.2 Treatment at a regional or tertiary hospital can be a burden for rural patients.3 Tertiary care hospitals are often far from home, and family members have increased expenses and stressors related to visiting ill loved ones.4,5

With tertiary care centers near capacity and smaller community hospitals sitting nearly empty,6 some have wondered how regionalization can better utilize existing healthcare resources in financially strained health systems at both ends of the hospital-size spectrum. One proposed strategy to better align healthcare resources with utilization is the concept of “back-transfer”—transferring patients near the end of their acute hospitalization to a lower-acuity hospital for the completion of their care.6 This is part of healthcare regionalization in that patient needs are matched to hospital capabilities, but in this case, those needs can be met at a less-specialized hospital.6

Back-transfer has been described previously for select patient populations. In Europe and Canada, the practice is used for cardiac patients after uncomplicated angiography and percutaneous coronary intervention (PCI) in specialty centers.7-10 Matteau et al suggest immediate back-transfer would be safe for up to two-thirds of ST-elevation myocardial infarction patients after PCI.10 In the United States, back-transfer (or “back-transport”) is used for the continued care of neonates after treatment at a tertiary neonatal intensive care unit.11,12 However, for adult populations in the United States, this concept has not been widely adopted and application of this transfer practice has not been studied.

In this study, we describe a cohort of patients transferred from larger hospitals to smaller hospitals for ongoing acute care over a 9-year period in a largely rural midwestern state. We sought to identify factors contributing to back-transfer, with the goal of elucidating the current use of back-transfer. We also discuss systematic barriers and potential benefits of its more widespread adoption.

METHODS

Study Design and Participants

This was an observational unmatched case-control study of all adults (18 years or older) hospitalized in Iowa between 2005 and 2013. Patients were identified from administrative claims data in the Iowa Hospital Association inpatient data set. Patient records were linked across inter-hospital transfer using a probabilistic linkage algorithm incorporating date of birth, sex, patient zip code, county of residence, visit date, and Social Security number through sequential matching. Records were limited to those eligible for back-transfer. Eligibility was defined as at least a 3-day inpatient hospitalization, residence at least 32 km from a tertiary center and within 32 km of a nontertiary hospital. The University of Iowa institutional review board approved the project under waiver of informed consent. This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.13

Measurements

Rurality was defined based on zip code of residence according to Rural-Urban Commuting Area (RUCA) definitions.14 Geographic distances were calculated by the Google Maps Application Programming Interface15 as the driving distance between each hospital and the centroid of the zip code of residence. Diagnosis categories were derived from International Classification of Diseases, 9th Edition, Clinical Modification (ICD-9-CM) codes categorized using the Clinical Classifications Software (CCS) available from the Agency for Healthcare Research and Quality Health Care Utilization Project (HCUP).16 Comorbidities were defined using Elixhauser methodology.17 Surgical procedures were defined as any invasive or noninvasive procedure typically performed in a surgical suite, identified using the Surgery Flag Software available from HCUP.18 Cardiac catheterization was defined by these ICD-9-CM procedure codes: 3722, 3723, 8850, 8851, 8852, 8853, 8854, 8855, 8856, 8857, 8858, 3606, 3607, 3609, and 0066. CAHs were defined by federal designation at the time of transfer.

Identification of Back-Transfers

Back-transfer was defined as a transfer to a smaller, less-specialized hospital as part of an acute care hospitalization. An algorithm using both emergency department (ED) volume and state trauma-level designation was used to identify back-transfers. First, hospitals were rank-ordered by annual ED census. Then, hospital trauma level was examined, and higher-level hospitals (eg, state trauma level II hospitals) ranked below lower-level hospitals (eg, state trauma level III hospitals) were moved on the list so the trauma level designation was preserved. Back-transfers were patients who were transferred from a larger hospital to a smaller one, requiring an interval of at least 5 steps down the rank-ordered list. The 5-step threshold was selected after careful examination by 2 independent investigators. Hospitals within 5 steps were felt to be indistinguishable in medical capabilities. For the purposes of this manuscript, the “large” hospital is defined as having a higher annual ED volume and/or trauma designation, and the “small” hospital as having lower annual ED volume and/or trauma designation. Once back-transfers were identified, the investigators observed that all back-transfers originated from 5 large Iowa hospitals. The data set was further limited to only those 5 hospitals. Demographics, diagnoses, procedures, insurance status, rurality of patient residence, and geographic variables were collected.

Analysis

Univariate analyses were performed using the t test, c2 test, or Wilcoxon rank-sum test, as appropriate. For the analysis, primary diagnosis is a mutually exclusive category, as an individual can have only 1 primary diagnosis. On the other hand, individuals can have multiple comorbidities. Thus, for the primary diagnosis, group-wise analyses were performed, and for the comorbidities, individual comparisons were made. A multivariable logistic regression explanatory model was developed to identify factors independently associated with back-transfer. Candidate variables for inclusion were selected if univariate analysis suggested a relationship with the dependent variable (back-transfer; P <.20). Final model variables were selected based on Bayesian Information Criterion to achieve a parsimonious final model. Interactions and co-linearity were tested in the model, but none were significant. Statistical analysis was conducted using SAS version 9.4 (SAS Institute; Cary, North Carolina) and Stata version 13.1 (StataCorp; College Station, Texas).

RESULTS

Over the 9-year period, 172,544 patients meeting inclusion criteria were admitted, of which 287 (0.2%) patients were back-transferred (Figure). All back-transfers were initiated by 5 large hospitals and accepted by 39 smaller Iowa hospitals, of which 16 were CAHs. There was no significant change in hospital admission volume or in back-transfers over the period of this study.

Univariate Analysis

Demographic comparison is shown in Table 1. Back-transferred patients were more likely than their nontransferred counterparts to be male, older, and white; to live in large rural areas; and to have public insurance.

Length of stay (LOS) differed by back-transfer status. Back-transferred patients were in the hospital longer than the control patients (median of 15 days vs 5 days; P <.001), and the median admission LOS prior to back-transfer was 8 days (Table 2). Total LOS for back-transferred patients was negatively related to rurality (measured by RUCA). As rurality increased, LOS decreased, with a median LOS of 17 days estimated among urban residents compared with 10 days for residents of small rural areas (P = .009). This trend was not found in the control group, which had a median LOS of 5 days regardless of rurality of residence. Those transferred to CAHs had shorter median posttransfer LOS than those transferred to non-CAHs (4.7 days vs 8.7 days; P = .001).

In most cases, patients were not hospitalized for acute care at the nearest hospital to their home (Table 2). Patients who were back-transferred were initially admitted to hospitals much farther from home (143 km vs 87 km; P <.0001). After back-transfer, patients were much closer to home (median = 10 km) and 53% were sent to the hospital nearest their home. Only 8% (n = 22) of back-transferred patients were transferred to CAHs.

Table 3 shows the disease characteristics of the study population. Controls and back-transferred patients had similar prevalence of CCS level 1 diagnostic codes. The most common primary disease categories in both patient groups were neoplastic, cardiovascular, digestive disease, and injury/poisoning. Significant differences between the 2 groups were present (P <.0001), with the greatest differences in obstetric and musculoskeletal diseases. Back-transferred patients were more likely to have each of the comorbidities evaluated. Finally, univariate modeling showed surgical rates were similar between the 2 groups, but back-transferred patients were more likely to have a cardiac catheterization procedure (P <.0001).

Multivariable Modeling

 
Copyright AJMC 2006-2017 Clinical Care Targeted Communications Group, LLC. All Rights Reserved.
x
Welcome the the new and improved AJMC.com, the premier managed market network. Tell us about yourself so that we can serve you better.
Sign Up
×

Sign In

Not a member? Sign up now!