Effectiveness of Transitional Care Nursing Programs in Reducing Hospital Readmissions Among Patients with Chronic Diseases

Introduction

Hospital readmission within 30 days of discharge remains one of the most closely monitored indicators of care quality across health systems, with particular attention directed at patients discharged following exacerbations of chronic conditions such as heart failure and chronic obstructive pulmonary disease, both of which carry well-documented readmission rates frequently exceeding 20% within 30 days (Jencks et al., 2009). Readmissions occurring within this early post-discharge window are frequently attributed, at least in part, to preventable gaps in the transition from hospital to home, including incomplete medication reconciliation, inadequate patient understanding of self-management instructions, and delayed recognition of early warning signs of clinical deterioration in the outpatient setting (Coleman et al., 2006).

Nurse-led transitional care programs have emerged as one of the most extensively studied strategies for addressing this preventable readmission risk. Although program design varies considerably across implementations, most nurse-led transitional care models share a common structure, incorporating a structured home visit conducted by a registered nurse within the first several days following discharge, proactive telephone follow-up at defined intervals during the vulnerable post-discharge period, and dedicated nursing time devoted to medication reconciliation and reinforcement of self-management instructions specific to the patient’s discharge diagnosis (Naylor et al., 2004). The Transitional Care Model, developed by Naylor and colleagues, and the Care Transitions Intervention, developed by Coleman and colleagues, represent two of the most widely studied and disseminated nurse-led transitional care frameworks, both of which have demonstrated favorable effects on readmission in randomized trial settings (Coleman et al., 2006; Naylor et al., 2004).

Despite this favorable randomized trial evidence, comparatively fewer studies have evaluated nurse-led transitional care program effectiveness using real-world, retrospective observational data reflecting routine clinical implementation outside the more tightly controlled conditions of a randomized trial, and fewer still have applied rigorous quasi-experimental methods, such as propensity score matching, capable of adequately addressing the substantial baseline differences in illness severity and prior utilization that typically exist between patients who are and are not enrolled in a transitional care program under real-world referral patterns (Hansen et al., 2011). Without adequate adjustment for this confounding by indication, in which sicker or higher-risk patients may be preferentially referred to transitional care programs specifically because of their elevated readmission risk, observational evaluations risk substantially underestimating the true program effect. The purpose of this study was to apply propensity score matching to a real-world retrospective cohort to evaluate the effect of enrollment in a nurse-led transitional care program on 30-day and 90-day all-cause readmission, time to readmission, and 30-day emergency department utilization among patients discharged following a chronic disease exacerbation.

Methods

Design and data source. This study used a retrospective cohort design with propensity score matching, drawing on electronic health record and administrative claims data from a single regional health system encompassing four acute care hospitals. The study cohort included adults discharged following an index hospitalization for heart failure, chronic obstructive pulmonary disease exacerbation, or a qualifying diagnosis among a pre-specified list of additional chronic conditions, including diabetes-related admission and chronic kidney disease exacerbation, between January 2022 and December 2024.

Exposure. The exposure of interest was enrollment in the health system’s nurse-led transitional care program, offered to eligible patients at the discretion of inpatient discharge planning staff based on standard eligibility criteria including qualifying diagnosis and absence of discharge to a skilled nursing or other post-acute facility. The program included a registered nurse home visit within 72 hours of discharge, incorporating medication reconciliation, vital sign assessment, and reinforcement of discharge self-management instructions; structured telephone follow-up calls at one and two weeks post-discharge; and a dedicated transitional care nurse serving as a single point of contact for the patient during the 30-day post-discharge period. Patients discharged during the same period who were eligible for but not enrolled in the program, most commonly due to program capacity constraints or patient decline rather than clinical exclusion, served as the usual-care comparison group.

Propensity score matching. A propensity score representing the predicted probability of transitional care program enrollment was estimated using logistic regression incorporating age, sex, qualifying diagnosis, Charlson Comorbidity Index, number of hospitalizations in the preceding 12 months, number of concurrent medications at discharge, insurance type, and whether the patient lived alone. Transitional care program participants were matched 1:1 to usual-care patients using nearest-neighbor matching without replacement within a caliper of 0.2 standard deviations of the logit of the propensity score. Covariate balance after matching was assessed using standardized mean differences, with a value below 0.10 considered indicative of adequate balance.

Outcome measures. The primary outcome was 30-day all-cause hospital readmission to any facility within the health system. Secondary outcomes included 90-day all-cause readmission, time to first readmission within 90 days, and 30-day emergency department visits not resulting in admission. Outcome data were obtained from the health system’s integrated electronic health record and supplemented with regional health information exchange data to capture readmissions occurring at non-system facilities.

Statistical analysis. Within the matched cohort, differences in binary outcomes were compared using McNemar’s test for paired proportions. Time to first readmission within 90 days was analyzed using Kaplan-Meier estimation and compared between groups using the log-rank test; adjusted hazard ratios were estimated using Cox proportional hazards regression stratified by matched pair, with additional adjustment for any residual covariate imbalance identified after matching. The proportional hazards assumption was assessed using Schoenfeld residuals. A two-sided p value of less than .05 was considered statistically significant.

Results

A total of 2,104 patients enrolled in the nurse-led transitional care program and 5,932 usual-care patients discharged during the study period were identified as potentially eligible for matching. Prior to matching, transitional care participants differed meaningfully from usual-care patients across several covariates, most notably a higher mean Charlson Comorbidity Index (4.3 versus 3.6, standardized mean difference 0.34) and a higher mean number of prior 12-month hospitalizations (2.1 versus 1.4, standardized mean difference 0.41), consistent with the expected pattern of confounding by indication in which higher-risk patients were preferentially referred to the program. Following 1:1 nearest-neighbor propensity score matching, 1,486 transitional care participants were matched to 1,486 usual-care patients, achieving substantial improvement in covariate balance, as summarized in Table 1 and illustrated in Figure 1.

As illustrated in Figure 1, matching substantially reduced imbalance across all examined covariates, with the largest pre-match imbalance observed for prior 12-month hospitalizations and Charlson Comorbidity Index, both of which were reduced to standardized mean differences below 0.05 following matching. This successful balancing supports the validity of the matched-cohort comparison for estimating the effect of transitional care program enrollment on subsequent readmission risk, independent of the substantial baseline confounding by indication present in the unmatched cohort.

Within the matched cohort of 1,486 pairs, 30-day all-cause readmission occurred in 132 of 1,486 transitional care participants (8.9%) compared with 250 of 1,486 usual-care patients (16.8%), a difference that was highly statistically significant (McNemar’s p < .001). Ninety-day readmission occurred in 241 transitional care participants (16.2%) compared with 428 usual-care patients (28.8%, p < .001). Thirty-day emergency department visits not resulting in admission occurred in 9.8% of transitional care participants compared with 15.4% of usual-care patients (p < .001).

As illustrated in Figure 2, the Kaplan-Meier curves for the two groups diverged early in the post-discharge period and continued to separate progressively through the full 90-day observation window, with the log-rank test confirming a highly significant difference in readmission-free survival between groups (p < .001). The curves showed the steepest divergence during the first two weeks following discharge, consistent with the timing of the transitional care program’s home visit and initial telephone follow-up contacts, both of which occur within this early, highest-risk window.

As summarized in Table 2, the adjusted Cox model estimated a 50% relative reduction in 30-day readmission hazard (adjusted hazard ratio 0.50, 95% CI 0.42–0.60, p < .001) and a 47% relative reduction in 90-day readmission hazard (adjusted hazard ratio 0.53, 95% CI 0.46–0.61, p < .001) among transitional care participants relative to matched usual-care patients. Thirty-day emergency department utilization was also significantly reduced (adjusted hazard ratio 0.61, 95% CI 0.52–0.72, p < .001). Ninety-day all-cause mortality was numerically lower among transitional care participants (3.1% versus 4.0%) but did not reach statistical significance (adjusted hazard ratio 0.77, 95% CI 0.56–1.06, p = .108), consistent with the relatively low absolute event rate for this outcome within the study’s follow-up window and matched sample size. The proportional hazards assumption was not violated for any outcome based on Schoenfeld residual testing.

Subgroup analysis by qualifying diagnosis indicated a directionally consistent readmission benefit across heart failure, chronic obstructive pulmonary disease, and other qualifying diagnoses, with the largest absolute risk reduction observed among the heart failure subgroup (30-day readmission 10.2% versus 21.4%, absolute risk reduction 11.2 percentage points), plausibly reflecting the higher baseline readmission risk characteristic of this population and correspondingly greater absolute room for improvement.

Discussion

This propensity-score matched retrospective cohort study found that enrollment in a nurse-led transitional care program was associated with substantial and statistically robust reductions in 30-day and 90-day all-cause readmission and 30-day emergency department utilization among patients discharged following a chronic disease exacerbation. The magnitude of the observed effect, an approximately 50% relative reduction in both 30-day and 90-day readmission hazard, is consistent with, and at the favorable end of, effect sizes reported in the foundational randomized trials of nurse-led transitional care models (Coleman et al., 2006; Naylor et al., 2004), providing valuable real-world confirmation that these program effects can be replicated under routine clinical implementation conditions outside the more controlled setting of a randomized efficacy trial.

The substantial pre-match imbalance identified between transitional care participants and usual-care patients, particularly with respect to comorbidity burden and prior hospitalization frequency, underscores the critical methodological importance of rigorous confounding adjustment when evaluating real-world transitional care program effectiveness using observational data. Had this analysis relied on a simple unadjusted comparison between enrolled and non-enrolled patients, the substantially higher baseline risk profile of transitional care participants, reflecting the health system’s evidently effective clinical practice of preferentially referring higher-risk patients to the program, would have been expected to bias the observed association toward the null or even toward an apparently harmful direction, potentially obscuring what this study’s propensity-matched analysis reveals to be a genuinely substantial protective effect (Hansen et al., 2011).

The pattern of early and progressive divergence in the Kaplan-Meier survival curves illustrated in Figure 2, with the steepest separation occurring during the first two weeks following discharge, aligns closely with the specific timing of the transitional care program’s most intensive contact points, including the home visit within 72 hours of discharge and the first telephone follow-up call at one week. This temporal correspondence supports a plausible causal mechanism in which early, structured nursing contact during the highest-risk post-discharge window enables timely identification and correction of medication discrepancies, early recognition of emerging clinical deterioration, and reinforcement of self-management instructions before a preventable complication progresses to the point of requiring acute care (Naylor et al., 2004).

The absence of a statistically significant mortality benefit, despite a favorable point estimate, should not be interpreted as evidence against the clinical value of this transitional care program, but rather reflects the comparatively low absolute mortality event rate within the 90-day follow-up window and the correspondingly limited statistical power available to detect a mortality difference of the modest magnitude suggested by the point estimate. Readmission and emergency department utilization, the primary outcomes for which this study was adequately powered, arguably represent the most direct and clinically actionable indicators of transitional care effectiveness, given that the specific care components comprising this program, medication reconciliation, symptom monitoring, and self-management reinforcement, are mechanistically most proximate to preventing an acute care encounter rather than directly modifying longer-term mortality risk.

Several limitations should be considered when interpreting these findings. Despite the substantial improvement in measured covariate balance achieved through propensity score matching, residual confounding by unmeasured factors, including functional status, caregiver support availability, and patient motivation or engagement with self-management, factors that may have independently influenced both the likelihood of transitional care program enrollment and readmission risk, cannot be fully excluded given the retrospective, non-randomized nature of this study design. This study was conducted within a single regional health system, and while readmission data were supplemented with regional health information exchange records to capture cross-facility readmissions, some readmissions occurring at facilities entirely outside the region may not have been captured, potentially resulting in modest underestimation of the true readmission rate in both groups. Finally, this study examined a program with a specific structure, including a 72-hour home visit and biweekly telephone follow-up; the magnitude of benefit observed here may not generalize directly to transitional care programs implemented with different staffing intensity, home visit timing, or follow-up duration.

Future research should extend this propensity-matched approach to additional health systems and geographic regions to assess the generalizability of the substantial effect sizes observed in this single-system analysis. Further investigation using instrumental variable or other quasi-experimental approaches capable of addressing residual unmeasured confounding, such as variation in program capacity constraints that create quasi-random variation in enrollment independent of patient risk, would help strengthen causal inference beyond what propensity score matching alone can provide. Additional research examining the relative contribution of specific program components, including the home visit, telephone follow-up, and medication reconciliation, to overall program effect would help health systems optimize resource allocation when scaling transitional care services under staffing constraints. Taken together, the findings of this study provide robust, real-world confirmatory evidence that nurse-led transitional care programs produce substantial, clinically meaningful reductions in preventable post-discharge hospital readmission and emergency department utilization among patients with chronic disease, supporting continued investment in and expansion of these programs within routine hospital discharge planning practice.

References

Austin, P. C. (2011). An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behavioral Research, 46(3), 399–424.

Coleman, E. A., Parry, C., Chalmers, S., & Min, S. J. (2006). The care transitions intervention: Results of a randomized controlled trial. Archives of Internal Medicine, 166(17), 1822–1828.

Hansen, L. O., Young, R. S., Hinami, K., Leung, A., & Williams, M. V. (2011). Interventions to reduce 30-day rehospitalization: A systematic review. Annals of Internal Medicine, 155(8), 520–528.

Jack, B. W., Chetty, V. K., Anthony, D., Greenwald, J. L., Sanchez, G. M., Johnson, A. E., Forsythe, S. R., O’Donnell, J. K., Paasche-Orlow, M. K., Manasseh, C., Martin, S., & Culpepper, L. (2009). A reengineered hospital discharge program to decrease rehospitalization: A randomized trial. Annals of Internal Medicine, 150(3), 178–187.

Jencks, S. F., Williams, M. V., & Coleman, E. A. (2009). Rehospitalizations among patients in the Medicare fee-for-service program. New England Journal of Medicine, 360(14), 1418–1428.

Naylor, M. D., Brooten, D. A., Campbell, R. L., Maislin, G., McCauley, K. M., & Schwartz, J. S. (2004). Transitional care of older adults hospitalized with heart failure: A randomized, controlled trial. Journal of the American Geriatrics Society, 52(5), 675–684.

Rennke, S., Nguyen, O. K., Shoeb, M. H., Magan, Y., Wachter, R. M., & Ranji, S. R. (2013). Hospital-initiated transitional care interventions as a patient safety strategy: A systematic review. Annals of Internal Medicine, 158(5_Part_2), 433–440.

Rosenbaum, P. R., & Rubin, D. B. (1983). The central role of the propensity score in observational studies for causal effects. Biometrika, 70(1), 41–55.

Verhaegh, K. J., MacNeil-Vroomen, J. L., Eslami, S., Geerlings, S. E., de Rooij, S. E., & Buurman, B. M. (2014). Transitional care interventions prevent hospital readmissions for adults with chronic illnesses. Health Affairs, 33(9), 1531–1539.

Voss, R., Gardner, R., Baier, R., Butterfield, K., Lehrman, S., & Gravenstein, S. (2011). The care transitions intervention: Translating from efficacy to effectiveness. Archives of Internal Medicine, 171(14), 1232–1237.

◆