Cameron J Gettel, MD, Department of Emergency Medicine, Brown University
55 Claverick Street, Providence, RI 02903
Phone: (717) 228-7759 Email: email@example.com
This paper has not been submitted or presented at a meeting. Dr Goldberg received support from the National Institute on Aging (R03AG056349) and the Agency for Healthcare Research and Quality (T32AG023482). The funders had no role in the design or conduct of the study; collection, management, analysis, and interpretation of the data; and preparation or approval of the manuscript.
1Department of Emergency Medicine, Brown University, Providence, RI
2Warren Alpert Medical School of Brown University, Providence, RI
We thank librarian Erika Sevetson for providing guidance with conducting literature searches. We thank Dilum Aluthge (DA) for screening all retrieved titles and abstracts.
This study aimed to identify interventions that are effective in improving the transitions of care for patients from nursing homes (NHs) to emergency departments (EDs). A total of 607 studies were identified, from which 19 studies were included for full-text review. Nine pre-post intervention studies and two retrospective cohort studies met all criteria for inclusion. In the quality assessment, two (18.2%) were assessed as good quality; seven (63.6%) were fair; and two (18.2%) were poor. Nine studies (81.2%) had a severe risk of bias, primarily due to confounding and deviation from the intended intervention. Pre-post intervention studies utilized transfer checklists/forms, web-based communication networks, and multimodal approaches to improve transitions of care. Eight studies reported significant improvement in critical NH-ED transfer information completeness after intervention implementation. Three studies assessed health care utilization after intervention implementation with two studies reporting no reduction in utilization and one study reporting decreased 30-day hospital readmission and ED revisit rates. Studies evaluating patient-centered outcomes, such as whether interventions reduced harm to patients by decreasing medical errors, hospital length of stay, or the overall number of facility transfers, are needed.
Key words: nursing home, emergency department, transitions of care, transfer form, health information technology
By 2030, 61 million people in the United States will be older than 65 years, and they will account for approximately 23% of the population.1,2 The number of people utilizing nursing home (NH) facilities, residential care places, or home care services is anticipated to reach 27 million by 2050.3 Visits to the emergency department (ED) by older patients are a driving force in emergency care demand, as 1.6 ED visits occur annually for each NH resident.4,5 Recent payment reforms, including bundled payments and accountable care organizations, have also attempted to incentivize NHs to prevent potentially avoidable hospitalizations and ED visits. Transitions of care occur with the movement of a patient from one health services setting to another. Information gaps conflicting care plans, duplicative testing, higher costs, medical errors, or delays in care.9-11 Frequent transitions of care are deemed undesirable due to the disorientation to the patient, thus this occurrence serves as a marker of poor provision of medical care in the long-term care (LTC) facility.
Despite the high prevalence of ED visits by NH residents and projected increases in the population of older adults aged 65 and older, little work has focused on how to improve NH-ED transitions of care. National geriatric and emergency medicine organizations, including the American Geriatrics Society and the American College of Emergency Physicians, proposed a minimal set of information for an “ideal transition record” and suggest communities develop standard data transfer forms.12,13 Prior efforts to improve transitions of care have included telephone conversations between staff, standardized communication tools, computer-based networks, and hospital-wide, patient-tracking dashboard systems.14-16 However, it is unclear which of these interventions are the most effective in improving patient-oriented and health services utilization outcomes.17
Given the continued increase in NH-ED transitions of care and the known importance of high-quality hand-offs to mitigate harm to individuals, interventions to standardize and improve NH-ED transitions of care are a population health priority. For this review, our aim was (1) to summarize the literature involving transitions of care for patients from NH facilities to EDs, and (2) to identify interventions that are effective in improving patient-oriented, systems-oriented, or health services utilization outcomes when a transfer occurs.
This was a systematic review of studies reporting the effectiveness of interventions targeting transitions of care from NHs to EDs. An a priori study protocol was registered in PROSPERO (CRD42018097222) that defined the objectives, search strategy, eligibility criteria, outcomes, and methods for data abstraction. This systematic review conforms to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The Population, Intervention, Comparison, Outcome (PICO) framework utilized in this systematic review can be seen in Table 1. We conducted a qualitative systematic review rather than a quantitative meta-analysis due to the heterogeneity of study designs, outcomes, and interventions. Institutional review board approval was not required because this study did not enroll patients or abstract data from medical records.
Studies were included if they were experimental and evaluated interventions targeting the NH-ED transition of care. Eligible study designs included randomized controlled trials, prospective observational studies, retrospective cohort studies, and before-after studies. Studies and research abstracts were included if participants experienced an NH-ED transition and received an intervention (eg, transfer checklist, health information technology) intended to improve the transition. For inclusion, studies were required to report on patient-oriented and health services utilization outcomes, including hospital use and ED revisits, or systems-oriented outcomes, including completeness of essential transfer correspondence information. We only included studies in which intervention participants were compared to a control group or outcomes were assessed in a pre-intervention vs post-intervention fashion. We excluded studies that included participants from assisted living facilities, studies that reported only on attitudes of providers and patients toward the transition, and non-English studies.
Together with a medical librarian we developed a database search strategy. We used the following databases: MEDLINE, Excerpta Medica dataBASE (EMBASE), Cumulative Index for Nursing and Allied Health (CINAHL), Cochrane Database of Systematic Reviews, and ClinicalTrials.gov. The MEDLINE search used a combination of key words and concepts including, but not limited to, “transitions of care,” “communication,” “patient transfer,” “transfer correspondence,” “geriatric,” “nursing home,” “skilled nursing facilities,” “emergency department,” and “emergency medicine.” These terms were translated to equivalent standardized terms for all database searches. Searches were completed in May 2018, and no limits were placed on publication date. The complete search strategies are listed in Supplementary Table 1 (All Supplementary items are available at the end of this article following the references or in the attached PDF). We also conducted a gray literature search using Web of Science, Google Scholar, and hand searches of the most recent abstracts in Academic Emergency Medicine, Annals of Emergency Medicine, and Journal of the American Geriatrics Society. All results were exported to EndNote.18
Two reviewers (NP, DA) independently screened all retrieved titles and abstracts, according to the inclusion and exclusion criteria. The level of agreement between reviewers to determine if studies should be included was quantified using a kappa analysis. Articles meeting inclusion underwent a full-text review. We searched the bibliographies of the included articles to find more potentially relevant studies that were not found in our original search. Questions regarding study inclusion were adjudicated by discussion among the study’s primary authors (CG, EG). We abstracted the key characteristics of the included manuscripts into a standardized data collection sheet, including the year, study design, practice setting, inclusion criteria, intervention, outcome assessment, and results.
Two reviewers (NP, CG) used the National Heart, Lung, and Blood Institute (NHLBI) Quality Assessment Tool for Before-After (Pre-Post) Studies with No Control Group to evaluate the overall quality of evidence for the identified before-after studies. The Quality Assessment Tool is comprised of 14 questions that qualitatively assess the research question, study population, eligibility criteria, sample size justification, exposure assessment, timeframe, outcome measures, blinding, follow-up rate, and statistical analyses. Two reviewers (NP, CG) used the Newcastle-Ottawa Quality Assessment Form for Cohort Studies to assess the overall quality for the included retrospective cohort studies. The Quality Assessment Form is comprised of a total of eight items within three groups: selection, comparability, and outcome. Any discrepancies were resolved by consensus.
Risk of Bias Assessment
Two reviewers (NP, CG) used the Cochrane Bias Methods Group and the Cochrane Non-Randomized Studies Methods Group Risk of Bias in Non-Randomized Studies – of Interventions (ROBINS-I) tool to evaluate for bias in the identified studies.19 The ROBINS-I tool assesses seven domains of bias, with classification of judgments into “low,” “moderate,” “serious,” and “critical” risk of bias. A “low risk of bias” domain is comparable to a well-performed randomized trial with regard to a specific domain. A “moderate risk of bias” domain is sufficient for a nonrandomized study with regard to a specific domain but cannot be considered comparable to a well-performed randomized trial. A “serious risk of bias” domain contains sizable problems. A “critical risk of bias” domain is too methodologically flawed to provide any useful evidence on the effects of the intervention. To be considered a “low risk of bias” study, the study must be judged to be at “low risk of bias” for all individual domains. To be considered a “critical risk of bias” study, the study must be judged to be at “critical risk of bias” in at least one domain.
Literature Flow and Included Studies
We identified 746 abstracts in the literature search. After duplicates were removed, 607 abstracts remained. From those, 588 studies were excluded on title and abstract review. We retrieved and fully reviewed the 19 remaining studies (Figure 1). The kappa value for interobserver agreement for study inclusion by reviewers was 0.942, indicating excellent agreement. We ultimately included a total of 11 studies: nine before-after studies and two retrospective cohort studies. These studies were published between 2005 and 2017. Authors of the primary studies were mainly from the United States. Other study authors were from Taiwan and Canada (Table 2 and Supplementary Table 2).
The total number of participants per study ranged from 74 to 843.20,21 The overall mean age of participants, calculated from the five studies with available data, was 81.8 years. The overall total percentage of female participants, calculated by pooling the percentage of female population from these five studies, was 58%. The percentage of participants with a documented diagnosis of dementia or cognitive impairment was 31.6% after pooling the population demographics from the two studies with available data.
We used the tool from the NHLBI to assess the nine studies that used before-after experimental designs (Table 3 and Supplementary Table 3). Overall, one (11.1%) study was assessed as good quality with the remaining studies assessed as fair (66.6%) or poor (22.2%) quality. The reason for fair- or poor-quality rating included: intervention not universally applied to all available individuals, failure to report whether a sufficient sample size was available to detect an association, and lack of blinding of outcome assessors (Supplementary Figure 1).
We used the Newcastle-Ottawa Quality Assessment Form for Cohort Studies (Table 2 and Supplementary Table 4) to assess the two retrospective cohort studies for quality. One study was rated as good quality, and the remaining study was rated as fair.
Using the ROBINS-I assessment tool to assess bias, we found that the 11 studies had a moderate (18.2%) risk or severe (81.2%) risk of bias (Table 4). The components with the highest risk of bias included: confounding and deviation from intended intervention (Supplementary Figure 2).
Characteristics of Interventions
Four studies used a checklist or form as the intervention.20-23 Three studies utilized a web-based communication network to facilitate transitions and reported on the presence of critical NH-ED transfer information.24-26 Two studies used a multimodal intervention (eg, statewide collaborative, plan-do-study-act cycles, root cause analyses) to improve NH-ED transitions of care.27,28 The two retrospective cohort studies reported whether the presence of a dedicated NH-ED transfer form improved the transfer of critical NH-ED transfer information.7,29
Fidelity of the Intervention
Implementation of the study intervention varied, with the transfer form or checklist present 22.7% to 100% of the time. When addressed in the nine pre-post studies, four of the studies had less than 50% uptake of the intervention in the post-intervention period.22-24,26
Documentation of Critical NH-ED Transfer Information
Six pre-post studies and two retrospective studies assessed the completeness of transition documentation when compared to predetermined critical NH-ED transfer information. All eight studies determined that documentation of critical NH-ED transfer information increased after the intervention was initiated.7,20,22,24,26-29 Individual studies differed in the number and magnitude of improvements in documentation of critical NH-ED transfer information. None reported whether greater completeness reduced patient harm.
Health Care Utilization After Intervention Initiation
Three pre-post studies assessed the impact of an intervention on hospital readmission and ED revisit rates. Two studies21,23 determined that 30-day readmission rates and 30-day ED revisit rates did not improve after the intervention was initiated, whereas one study found a decrease in 30-day hospital readmission rates and 30-day ED revisit rates after introduction of a Universal Transfer Form.25 In comparison to the studies that did not find a difference in utilization rates, research staff from the latter study used an electronic clinical documentation tool. The clinical documentation tool allowed clinical information to be reflected in the state’s health information exchange (HIE). As a result, local ED clinicians were able to access this relevant patient information through the HIE.
In this systematic review of 11 studies, we found that efforts to improve NH-ED transitions of care have focused on transfer information content and whether the intervention improved health care utilization. While most interventions led to an improvement in documentation completeness by using checklists or standardized transfer forms, only one intervention resulted in decreased health care utilization.25 No studies reported whether interventions reduced patient harms.
To our knowledge, this is the first systematic review to synthesize the available literature on interventions to improve NH-ED transitions of care. We anticipate these results will be of interest to patients, clinicians, NH administrators, researchers, and policymakers who are working to identify and disseminate effective techniques to improve transitions of care.
We have several recommendations to improve the lack of internal validity and fidelity of the interventions in the currently available literature on this topic. First, the groups assessed in the included studies were frequently dissimilar. Without multivariable modeling to adjust for variation between the two groups, it is difficult to determine the independent effect of the intervention.30 Second, many of the pre-post studies included in this review did not deliver the intervention consistently, allowing for the introduction of selection bias. We believe nursing factors (limited nurse-resident ratio) and patient factors (cognitive impairment hindering patient health information transmission) may have reduced the fidelity of the interventions.7,21 More emphasis should be placed on creating interventions that are more easily adopted by NH staff. Researchers could learn from one included study21 that achieved consistent intervention fidelity by employing a comprehensive one-page NH-ED checklist. This checklist was developed and improved through an iterative process of stakeholder qualitative interviews and expert opinion.
Research and Clinical Implications
Developing effective interventions to aid NH-ED transitions of care will be particularly important as the number of older adults continues to rise and EDs provide an increasing amount of care to this population. With incentives changing to promote the coordination of care across health care settings, the Interventions to Reduce Acute Care Transfers (INTERACT) program has been developed to assist LTC staff to avoid hospitalizations by improving the management of acute conditions. The INTERACT program is based on three key tenets: (1) recognition and management of acute conditions before becoming severe enough to necessitate hospitalization; (2) providing communication, documentation, and decision support tools to allow effective management in the NH when safe and feasible; and (3) emphasizing advance care planning, hospice, and palliative care to encourage goals-of-care discussions and reduce hospitalizations in those with end-stage illnesses where the risks of hospitalization often outweigh the benefits.31 An earlier nonrandomized pilot study of INTERACT found a 24% reduction in all-cause hospitalizations in participating NHs, compared with a 6% reduction in those that did not implement INTERACT.32 A more recent randomized implementation trial assessing NHs utilizing INTERACT found a statistically nonsignificant reduction in hospitalization rates compared with control NHs.31 Once the decision is made to transfer care of the patient, clinicians caring for older persons in NHs and EDs should strive to share pertinent clinical information during transfers. This information should include the patient’s preferences, clinical status, medication reconciliation, updated contact information, and baseline physical and mental function. The INTERACT Version 4.0 Tool is a quality improvement tool geared toward NH staff to promote the effective communication of critical information to the ED staff.33
Future studies should report on patient-centered outcomes such as reducing medical errors, adverse drug events, prolonged hospitalization, duplicative testing, and the frequency of transfers. NH staff are instrumental in facilitating transitions and should be engaged to develop, create, and vet possible solutions. The studies in this systematic review focused on the transmission of critical NH-ED transfer information to the new site of care, however, future studies could also focus on what EDs can do to improve receipt of high-quality information. More emphasis should be given to web-based techniques and other technologies to improve NH-ED transitions of care. NHs and other LTC facilities have been slow to adopt health information technologies (HITs), with prior research revealing a lack of necessary technology support and infrastructure, underinvestment in NH staff training, and the absence of a systematic process for HIT implementation.34 Policymakers should consider addressing barriers by creating incentives to improve NH staff training and technology infrastructure.
This systematic review has several limitations. First, significant heterogeneity of the included interventions and outcomes prevented meta-analysis and reduced the ability to draw definitive and actionable conclusions. Our definition of intervention may have been too narrow, potentially missing studies that could have fit within a broader definition and included interventions initiated once the patient reached the ED. Many of the studies evaluated the care transition to a single ED, reducing the generalizability of the intervention. We limited our review to English-language publications and may have excluded relevant studies in other languages. Finally, it is possible that we failed to identify potentially relevant studies, but we employed best practices in conducting this systematic review including searching bibliographies of relevant articles and searching multiple databases.
In conclusion, we found that most studies assessing interventions used in NH-ED transitions of care targeted improved capture of critical NH-ED transfer information and very few examined subsequent health care utilization. While most studies achieved “improved” documentation, these studies were of fair to poor quality and had a severe risk of bias. Further studies are needed to evaluate what matters most to patients as they experience transitions in care. We anticipate these outcomes will be the number of transitions, medical errors, adverse drug events, and duplicative testing and procedures. This work is urgent as the population of older adults is increasing and efforts to curb medical errors and associated costs are necessary.
1. Knickman JR, Snell EK. The 2030 problem: caring for aging baby boomers. Health Serv Res. 2002;37(4):849-884. doi:10.1034/j.1600-0560.2002.56.x
2. Vincent GK, Velkoff VA. The next four decades - the older population in the United States: 2010 to 2050. Current Population Reports. https://www.census.gov/prod/2010pubs/p25-1138.pdf. Published May 2010. Accessed January 8, 2020.
3. US Department of Health and Human Services’ Office of the Assitant Secretary for Planning and Evaluation. The future supply of long-term care workers in relation to the aging baby boom generation: report to Congress. aspe.hhs.gov website. https://aspe.hhs.gov/pdf-report/future-supply-long-term-care-workers-relation-aging-baby-boom-generation. Published May 14, 2003. Accessed January 8, 2020.
4. Pallin DJ, Allen MB, Espinola JA, Camargo CA Jr, Bohan JS. Population aging and emergency departments: visits will not increase, lengths-of-stay and hospitalizations will. Health Aff (Millwood). 2013;32(7):1306-1312. doi:10.1377/hlthaff.2012.0951
5. Wang HE, Shah MN, Allman RM, Kilgore M. Emergency department visits by nursing home residents in the United States. J Am Geriatr Soc. 2011;59(10):1864-1872. doi:10.1111/j.1532-5415.2011.03587.x
6. Morphet J, Griffiths DL, Innes K, Crawford K, Crow S, Williams A. Shortfalls in residents’ transfer documentation: challenges for emergency department staff. Australas Emerg Nurs J. 2014;17(3):98-105. doi:10.1016/j.aenj.2014.03.004
7. Cwinn MA, Forster AJ, Cwinn AA, Hebert G, Calder L, Stiell IG. Prevalence of information gaps for seniors transferred from nursing homes to the emergency department. CJEM. 2009;11(5):462-471.
8. Gettel CJ, Merchant RC, Li Y, et al. The impact of incomplete nursing home transfer documentation on emergency department care. J Am Med Dir Assoc. 2018;20(8):935-941.e3. doi:10.1016/j.jamda.2018.09.008
9. Meisel ZF, Pollack CV. Patient safety in emergency care transitions. Emergency Med Reports. 2006. https://www.reliasmedia.com/articles/120828-patient-safety-in-emergency-care-transitions. Accessed January 8, 2020.
10. Bodenheimer T. Coordinating care--a perilous journey through the health care system. N Engl J Med. 2008;358(10):1064-1071. doi:10.1056/NEJMhpr0706165
11. Coleman EA, Parry C, Chalmers S, Min SJ. The care transitions intervention: results of a randomized controlled trial. Arch Intern Med. 2006;166(17):1822-1828. doi:10.1001/archinte.166.17.1822
12. Snow V, Beck D, Budnitz T, et al. Transitions of care consensus policy statement American College of Physicians-Society of General Internal Medicine-Society of Hospital Medicine-American Geriatrics Society-American College of Emergency Physicians-Society of Academic Emergency Medicine. J Gen Intern Med. 2009;24(8):971-976. doi:10.1007/s11606-009-0969-x
13. Terrell KM, Hustey FM, Hwang U, et al. Quality indicators for geriatric emergency care. Acad Emerg Med. 2009;16(5):441-449. doi:10.1111/j.1553-2712.2009.00382.x
14. Record JD, Niranjan-Azadi A, Christmas C, et al. Telephone calls to patients after discharge from the hospital: an important part of transitions of care. Med Educ Online. 2015;20:26701. doi:10.3402/meo.v20.26701
15. Hewner S, Sullivan SS, Yu G. Reducing emergency room visits and in-hospitalizations by implementing best practice for transitional care using innovative technology and big data. Worldviews Evid Based Nurs. 2018;15(3):170-177. doi:10.1111/wvn.12286
16. Ayele RA, Lawrence E, McCreight M, et al. Study protocol: improving the transition of care from a non-network hospital back to the patient’s medical home. BMC Health Serv Res. 2017;17(1):123. doi:10.1186/s12913-017-2048-z
17. Kessler C, Williams MC, Moustoukas JN, Pappas C. Transitions of care for the geriatric patient in the emergency department. Clin Geriatr Med. 2013;29(1):49-69. doi:10.1016/j.cger.2012.10.005
18. Clarivate Analytics. Endnote Version X8.2. 2018.
19. Sterne JA, Hernan MA, Reeves BC, et al. ROBINS-I: a tool for assessing risk of bias in non-randomized studies of interventions. BMJ. 2016;355:i4919. doi:10.1136/bmj.i4919
20. Kelly NA, Mahoney DF, Bonner A, O’Malley T. Use of a transitional minimum data set (TMDS) to improve communication between nursing home and emergency department providers. J Am Med Dir Assoc. 2012;13(1):85.e9-e15. doi:10.1016/j.jamda.2011.02.007
21. Tsai HH, Tsai YF. Development, validation and testing of a nursing home to emergency room transfer checklist. J Clin Nurs. 2018;27(1-2):115-122. doi:10.1111/jocn.13853
22. Terrell KM, Brizendine EJ, Bean WF, et al. An extended care facility-to-emergency department transfer form improves communication. Acad Emerg Med. 2005;12(2):114-118. doi:10.1197/j.aem.2004.10.013
23. Zafirau WJ, Snyder SS, Hazelett SE, Bansal A, McMahon S. Improving transitions of care: efficacy of a transfer form in communicating patients wishes. Am J Med Qual. 2012;27(4):291-296. doi:10.1177/1062860611427413
24. Hustey FM, Palmer RM. An internet-based communication network for information transfer during patient transitions from skilled nursing facility to the emergency department. J Am Geriatr Soc. 2010;58(6):1148-1152. doi:10.1111/j.1532-5415.2010.02864.x
25. Yeaman B, Ko KJ, Alvarez del Castillo R. Care transitions in long-term care and acute care: health information exchange and readmission rates. Online J Issues Nurs. 2015;20(3):5.
26. Zamora Z, McCall B, Patel L, et al. Implementation of a web-based system to improve the transitional care of older adults. J Nurs Care Qual. 2012;27(2):182-189. doi:10.1097/NCQ.0b013e318237af71
27. Carson J, Gottheil S, Gob A, Lawson S. London Transfer Project: improving handover documentation from long-term care homes to hospital emergency departments. BMJ Open Qual. 2017;6(2):e000024. doi:10.1136/bmjoq-2017-000024
28. Tupper JB, Gray CE, Pearson KB, Coburn AF. Safety of rural nursing home-to-emergency department transfers: improving communication and patient information sharing across settings.
J Healthc Qual. 2015;37(1):55-65. doi:10.1097/01.JHQ.0000460120.68190.15
29. Dalawari P, Duggan J, Vangimalla V, Paniagua M, Ambrecht ES. Patient transfer forms enhance key information between nursing homes and emergency department. Geriatr Nurs. 2011;32(4):270-275. doi:10.1016/j.gerinurse.2011.05.001
30. Broderick KB, Ranney ML, Vaca FE, et al. Study designs and evaluation models for emergency department public health research. Acad Emerg Med. 2009;16(11):1124-1131. doi:10.1111/j.1553-2712.2009.00557.x
31. Kane RL, Huckfeldt P, Tappen R, et al. Effects of an intervention to reduce hospitalizations from nursing homes: a randomized implementation trial of the INTERACT program. JAMA Intern Med. 2017;177(9):1257-1264. doi:10.1001/jamainternmed.2017.2657
32. Ouslander JG, Lamb G, Tappen R, et al. Interventions to reduce hospitalizations from nursing homes: evaluation of the INTERACT II collaborative quality improvement project. J Am Geriatr Soc. 2011;59(4):745-753. doi:10.1111/j.1532-5415.2011.03333.x
33. Med-Pass Inc. INTERACT Tools - INTERACT Tools - Interventions to Reduce Acute Care Transfers. med-pass.com website. www.med-pass.com/interact-4. Accessed January 8, 2020.
34. Ko M, Wagner L, Spetz J. Nursing home implementation of health information technology: review of the literature finds inadequate investment in preparation, infrastructure, and training. Inquiry. 2018;55:46958018778902. doi:10.1177/0046958018778902