Outpatient Follow-Up Visits to Reduce 30-Day All-Cause Readmissions for Heart Failure, COPD, Myocardial Infarction, and Stroke: A Systematic Review and Meta-Analysis
SYSTEMATIC REVIEW — Volume 21 — September 26, 2024
PEER REVIEWED
In the first step, 2,830 records were identified through database searching: 1,343 in PubMed, 754 in CINAHL, and 733 in Cochrane. After 574 duplicates were removed, the titles and abstracts of 2,256 studies were screened; of these, 2,224 studies were excluded. Thirty-two studies were assessed for eligibility. Of these, 18 were excluded for the following reasons: in 6 studies, we could not isolate effect; 4 studies had the wrong intervention; the sample size was too small in 3 studies; 3 studies required that a follow-up visit had to occur at only 1 specific outpatient clinic; 1 study had only an abstract, and 1 study described a study already included in the analysis. In addition, we included 1 study from the bibliographies of 7 related meta-analyses reviewed. In the final analysis, 15 studies were included: 14 from the full-text review and 1 from a related meta-analysis. Of these, 10 were included in the quantitative meta-analysis.
Figure 1.
PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analysis) flow diagram of systematic review of US studies of outpatient follow-up visits and reduction of 30-day all-cause readmissions among patients with heart failure, chronic obstructive pulmonary disease, acute myocardial infarction, or stroke. Abbreviation: CINAHL, Cumulative Index to Nursing and Allied Health Literature.
Figure 2.
Random effect meta-analysis showing the pooled summary estimate of all 10 Tier 1 studies. The size of the data markers (squares) corresponds to the weight of the study in the meta-analysis. Abbreviations: HR, hazard ratio; OR, odds ratio.
| Study | OR/HR (95% CI) | Weight, % |
|---|---|---|
| Baecker et al (25), 2020 | 0.88 (0.82–0.95) | 12.5 |
| Budde et al (23), 2019 | 1.14 (0.89–1.47) | 9.1 |
| Distelhorst and Hansen (26), 2022 | 0.70 (0.50–0.97) | 7.4 |
| Fidahussein et al (21), 2014 | 1.02 (0.79–1.31) | 9.1 |
| Hussein et al (30), 2022 | 0.30 (0.18–0.50) | 4.6 |
| Lee et al (24), 2016 | 0.85 (0.73–0.98) | 11.3 |
| Leppert et al (29), 2020 | 0.87 (0.76–1.00) | 11.5 |
| Sharif et al (22), 2014 | 0.70 (0.57–0.86) | 10.1 |
| Terman et al (28), 2018 | 0.98 (0.97–0.99) | 13.0 |
| Xu et al (27), 2022 | 0.52 (0.45–0.60) | 11.4 |
| Overall: I-squared = 92.7%; P <.001 | 0.79 (0.69–0.91) | 100 |
Figure 3.
Random effect meta-analysis showing the pooled effect estimates for 3 subgroups.
| Comparison | No of studies | OR/HR (95% CI) | P value |
|---|---|---|---|
| High quality vs low quality | |||
| High quality | 7 | 0.82 (0.71–0.95) | .008 |
| Low quality | 3 | 0.65 (0.37–1.15) | .14 |
| Controlled for time-dependent bias | |||
| Not controlled | 6 | 0.69 (0.50–0.95) | .02 |
| Controlled | 4 | 0.91 (0.83–0.99) | .02 |
| Grouped by disease | |||
| Chronic obstructive pulmonary disease | 3 | 0.93 (0.68–1.26) | .62 |
| Heart failure | 4 | 0.73 (0.55–0.95) | .02 |
| Stroke | 3 | 0.76 (0.57–1.01) | .06 |
| Overall | — | 0.79 (0.69–0.91) | .001 |