Abstract
Abstract
Objective
Unplanned hospital readmissions within 30 days of discharge measure the quality of healthcare. This study aims to identify the characteristics of patients at higher risk of readmission.
Design
Retrospective cohort study.
Setting
North-east Italy (Marca Trevigiana Local Health Authority).
Data source
The study examined a total of 39 467 index admissions from hospital discharges (SDO) in the 890 000-inhabitant area during 2022.
Outcome measure
Readmission rates and 95% CIs were computed by risk factor, age and type of admission (surgical or medical). A logistic mixed-effects model was used to estimate readmission OR, adjusting for potential confounders.
Results
A total of 2197 readmissions occurred within 30 days of the index admission, resulting in an overall rate of 30-day readmissions of 6.7% (CI 6.4% to 7.0%). The median time to readmission was 11 days (IQR 5 to 20). In the multivariate analysis, after adjusting for age and sex, the following clinical conditions were associated with a higher risk of readmission: alcohol-related disease (OR=2.06, CI 1.36 to 3.13), metastatic cancer (OR=1.98, CI 1.57 to 2.50), epilepsy (OR=1.93, CI 1.36 to 2.75), dialysis or end-stage kidney disease (OR=1.92, CI 1.39 to 2.66), chronic obstructive pulmonary disease (OR=1.88, CI 1.49 to 2.36), stoma (OR=1.72, CI 1.22 to 2.44), transplant (OR=1.62, CI 1.03 to 2.55), being bedridden (OR=1.57, CI 1.28 to 1.93), anaemia (OR=1.57, CI 1.35 to 1.83), urinary tract infection (OR=1.54, CI 1.30 to 1.83), pneumonia (OR=1.52, CI 1.31 to 1.75), dementia (OR=1.49, CI 1.24 to 1.79), diabetes (OR=1.37, CI 1.17 to 1.61) and transfusion (OR=1.34, CI 1.03 to 1.73).
Conclusion
Several chronic and acute conditions at index admission significantly increased the risk of readmission. Strengthening transitional care, outpatient services and palliative care could mitigate readmissions.
Keywords: readmission, transitional care, chronic conditions, healthcare policy, administrative records
STRENGTHS AND LIMITATIONS OF THIS STUDY.
The use of routinely collected hospital discharge records allowed the inclusion of a large, population-based cohort, reducing selection bias.
The availability of standardised International Classification of Diseases, Ninth Revision codes enabled consistent identification of comorbidities but limited clinical detail and disease severity assessment.
The mixed-effects logistic regression model accounted for clustering at the hospital level, strengthening the validity of the estimated associations.
The retrospective cohort design relies on administrative data, which may be subject to coding errors and misclassification of diagnoses and procedures.
Potential confounding from unmeasured variables (eg, socioeconomic status, outpatient care, functional status) could not be addressed due to data limitations.
Introduction
Hospital readmissions, particularly within 30 days of discharge, are critical to healthcare quality.1 2 High readmission rates negatively affect patient outcomes and result in substantial financial penalties for healthcare institutions. Preventing hospital readmissions remains a priority in both clinical practice and policymaking, aimed at enhancing the quality of care and minimising unnecessary healthcare expenses.3
In the US, the hospital readmissions reduction programme,1 4 5 which penalises hospitals for excessive readmissions for selected medical conditions under Medicare,5 was partly ineffective;6 7 readmission rates declined from 21.5% to 17.8% for targeted conditions and from 15.3% to 13.1% for non-targeted conditions.8 9
However, it is crucial to identify patients at higher risk of readmission to help develop global measures to address the phenomenon in the broader context of public health, using routinely collected data. Previous studies have shown that patients with multimorbidity, frailty and social vulnerability are at higher risk of readmission.36 7 9,13 Conversely, implementing a rehabilitation programme in home and nursing facility-based discharge settings has reduced the risk of hospital readmission.14 15
Research has identified various models for predicting hospital readmission risks.16 17 The majority of these models designate comorbidity, length of stay and prior hospital utilisation frequency as the most influential variables.16 17 However, many of these models exhibit only a modest to limited capacity to accurately predict 30-day all-cause readmissions, thereby restricting their practical application.16 17
This study aims to identify the characteristics of patients at higher risk of readmission within 30 days of discharge and to formulate recommendations for healthcare providers, social services and policymakers to mitigate the effects of this critical phenomenon from a continuous improvement perspective.
Methods
Study design and data source
This study uses a retrospective cohort design. The database used uses data from regular admissions collected in 2022 through the administrative electronic flow of hospital discharge records (SDO) for a population of 890 000 served by ULSS 2 Health Authority Marca Trevigiana in north-east Italy. Readmission was defined as an unplanned hospital admission within 30 days after discharge from the initial admission. Exclusion criteria included: (1) admissions of individuals under 18 years old and (2) Admissions ending with discharge to another hospital or patient death.
Data extracted
Risk factors for readmission that could be obtained from hospital discharge records and are summarised as follows:
Patient demographic characteristics (age, gender).
Admission characteristics (length of stay, hospital, type of admission - medical or surgical, type of discharge - home, nursing home, home care).
Patient conditions and comorbidities (dementia, alcohol abuse, epilepsy, obesity, bed confinement, diabetes, dialysis or advanced stage kidney disease, transplant, pressure injuries, COVID-19 positivity, chronic obstructive pulmonary disease (COPD), cancer, stoma, infections (coded as in table 1 using International Classification of Diseases, Ninth Revision (ICD-9) classification)).
Procedures occurring at index admission (blood transfusion).
Table 1. International Classification of Diseases, Ninth Revision codes for risk factors.
| Clinical risk factors | ICD 9 codes |
|---|---|
| Dementia | 290-294 |
| Alcohol-related disease | 5710-5712; 34831; 34982 |
| Bedridden | V4984; 7282; 7994; 2639; 707 |
| Diabetes | 250 |
| Dialysis or end-stage kidney disease | 5855-5856, V560-568; 3942, 3927; 3995 |
| Transplant | V420-V429 (excl. V4289) |
| Epilepsy | 345 |
| Obesity | 278 |
| Transfusion | 990-999 |
| Anaemia | 280-285 |
| COVID-19 | 48041; 04311; 04321; 51971 |
| Stoma | V44; V55 |
| COPD | 490-496 |
| Cancer | 140-209 |
| Myeloma | 20300 |
| Pneumonia | 480-486; 4870 |
| UTI | 599; 590XX; 5950 |
COPD, chronic obstructive pulmonary disease; ICD 9, International Classification of Diseases, Ninth Revision; UTI, urinary tract infection.
Statistical analysis
The difference in distribution for categorical variables was tested using the χ2 test, while the t-test was applied to verify the mean difference between groups for continuous variables.
The readmission rate was calculated as the number of readmissions divided by the number of index admissions. Non-overlapping CIs, calculated using the modified Wald method, were considered statistically significant differences.
Mixed-effects multiple logistic regression was used to assess the impact of individual predictors on the probability of readmission, accounting for hospital nesting and controlling for age and gender.
Statistical tests were two-sided and a p value <0.05 was considered statistically significant. All analyses were performed using R V.4.2.3 (R Foundation for Statistical Computing, Vienna, Austria).
Patient and public involvement
None.
Results
In 2022, there were 32 700 index admissions in the area. 2197 readmissions occurred within 30 days of the index admission. The overall 30-day readmission rate was 6.7% (CI 6.4 to 7.0). The median time to readmission was 11 days (IQR 5 to 20).
Table 2 summarises the readmission probability for patients based on various risk factors. Older patients (75+) had a significantly higher readmission rate (9.7%) compared with younger patients (18–74 years), who had a readmission rate of only 4.7% (p<0.001). Regarding index admission characteristics, the average length of stay for patients who were later readmitted was 12.16 days, which was significantly longer than the 8.48 days for those who were not readmitted (p<0.001). Patients discharged to home care had a much higher readmission rate (17.0%) compared with those discharged to their homes (5.7%) or to a nursing home (9.7%, p<0.001). There was no significant difference in readmission rates between patients discharged on a weekend (6.5%) and those discharged on a weekday (6.9%, p=0.164). Several clinical conditions showed significant associations with readmission status; in particular, dementia (13.3, p<0.001), diabetes (9.6, p<0.001), alcohol-related diseases (12.9, p<0.001), dialysis or end-stage kidney disease (13.3, p<0.001), epilepsy (12.9, p<0.001), COPD (13.0, p<0.001), pneumonia (10.6, p<0.001), urinary tract infections (UTIs) (12.4, p<0.001), metastatic cancer (12.2, p<0.001), anaemia (12.0, p<0.001), stoma (12.3, p<0.001), transplant (10.4, p=0.031), the necessity of transfusions (14.4, p<0.001) and bedridden status (14.0, p<0.001) exhibited significantly higher rates of readmission compared with the absence of these conditions. There was no significant difference in readmission rates for patients with COVID.
Table 2. Probability of readmission by demographic, administrative and clinical risk factors.
| Risk factor | Yes (n=2197) | Total (n=32 700) | P value |
|---|---|---|---|
| Sex | 0.462 | ||
| F | 6.8 | 16 288 | |
| M | 6.6 | 16 412 | |
| Age | <0.001 | ||
| 18–74 | 4.7 | 19 598 | |
| 75+ | 9.7 | 13 102 | |
| Discharge | <0.001 | ||
| Home | 5.7 | 27 991 | |
| Nursing home | 9.7 | 2660 | |
| Home care | 17.0 | 2049 | |
| Length of stay | <0.001 | ||
| Mean (SD) | 12.16 (10.26) | 8.73 (8.38) | |
| Median (range) | 10.00 (2.00, 98.00) | 6.00 (1.00, 100.00) | |
| Discharge_day | 0.164 | ||
| Weekday | 6.9 | 20 683 | |
| Weekend | 6.5 | 12 017 | |
| Dementia | <0.001 | ||
| 0 | 6.5 | 31 442 | |
| 1 | 13.3 | 1258 | |
| Alcohol-related disease | <0.001 | ||
| 0 | 6.7 | 32 498 | |
| 1 | 12.9 | 202 | |
| Bedridden | <0.001 | ||
| 0 | 6.5 | 31 711 | |
| 1 | 14.0 | 989 | |
| Diabetes | <0.001 | ||
| 0 | 6.5 | 30 689 | |
| 1 | 9.6 | 2011 | |
| Dialysis or end-stage KD | <0.001 | ||
| 0 | 6.6 | 32 353 | |
| 1 | 13.3 | 347 | |
| Transplant | 0.031 | ||
| 0 | 6.7 | 32 489 | |
| 1 | 10.4 | 211 | |
| Epilepsy | <0.001 | ||
| 0 | 6.7 | 32 413 | |
| 1 | 12.9 | 287 | |
| Obesity | 0.436 | ||
| 0 | 6.7 | 32 316 | |
| 1 | 5.7 | 384 | |
| Transfusion | <0.001 | ||
| 0 | 6.6 | 32 094 | |
| 1 | 14.4 | 606 | |
| Anaemia | <0.001 | ||
| 0 | 6.3 | 30 405 | |
| 1 | 12.0 | 2295 | |
| COVID | 0.119 | ||
| 0 | 6.7 | 30 788 | |
| 1 | 7.6 | 1912 | |
| Stoma | <0.001 | ||
| 0 | 6.7 | 32 382 | |
| 1 | 12.3 | 318 | |
| COPD | <0.001 | ||
| 0 | 6.6 | 32 007 | |
| 1 | 13.0 | 693 | |
| Cancer | <0.001 | ||
| 0 | 6.6 | 28 970 | |
| 1a. Primary | 6.9 | 3019 | |
| 1b. Metastatic | 12.2 | 711 | |
| Myeloma | 0.678 | ||
| 0 | 6.7 | 32 611 | |
| 1 | 5.6 | 89 | |
| Pneumonia | <0.001 | ||
| 0 | 6.3 | 29 794 | |
| 1 | 10.6 | 2906 | |
| UTI | <0.001 | ||
| 0 | 6.5 | 31 243 | |
| 1 | 12.4 | 1457 |
COPD, chronic obstructive pulmonary disease; KD, kidney disease; UTI, urinary tract infection.
Table 3 presents readmission rates by age group and index admission type (medical or surgical). Older patients exhibited higher readmission rates, particularly following medical index admission (11.1%). In contrast, younger patients undergoing surgery had the lowest readmission rate (2.6%). For patients discharged to home care, readmission rates were significantly elevated across all categories: age 18–74 medical (19.0%); age 75+ medical (17.0%); age 18–74 surgical (11.1%) and age 75+ surgical (16.7%). The bedridden or sarcopenic/malnourished group also showed a higher rate of readmission after medical admission in both age categories 18–74 (13.8%) and 75+ (14.2%). Additionally, patients with dementia had significantly greater readmission rates, especially for those aged 75+ in surgical admissions (21.1%).
Table 3. Readmission rates by age, type of first admission (surgical or medical) and risk factors.
| Variable | Age=18–74; type=M | Age=75+; type=M | Age=18–74; type=S | Age=75+; type=S | Total |
|---|---|---|---|---|---|
| Rate/100 (CI) (N) | Rate/100 (CI) (N) | Rate/100 (CI) (N) | Rate/100 (CI) (N) | Rate/100 (N) | |
| Sex – F | 8.7 (7.8 to 9.6) (317) | 10.9 (10.0 to 11.8) (548) | 2.4 (2.0 to 2.8) (131) | 5.5 (4.5 to 6.4) (115) | 6.8 (6.4 to 7.2) (1111) |
| Sex – M | 6.9 (6.2 to 7.6) (319) | 11.5 (10.5 to 12.4) (466) | 2.7 (2.3 to 3.1) (158) | 7.5 (6.3 to 8.7) (143) | 6.6 (6.2 to 7.0) (1086) |
| Discharge - home | 7.1 (6.5 to 7.7) (533) | 9.9 (9.1 to 10.6) (594) | 2.5 (2.2 to 2.8) (274) | 5.6 (4.8 to 6.4) (188) | 5.7 (5.4 to 5.9) (1589) |
| Discharge - nursing home | 9.8 (6.9 to 12.6) (41) | 10.9 (9.4 to 12.3) (185) | 2.8 (−0.3 to 5.9) (3) | 7.0 (4.6 to 9.4) (30) | 9.7* (8.6 to 10.9) (259) |
| Discharge - home care | 19.1* (14.8 to 23.3) (62) | 17.1* (15.1 to 19.1) (235) | 11.1 (5.2 to 17.0) (12) | 16.7* (12.0 to 21.4) (40) | 17.0* (15.4 to 18.7) (349) |
| Dementia | 12.3 (6.7 to 18.0) (16) | 13.1 (11.0 to 15.1) (139) | – | 21.1* (10.5 to 31.6) (12) | 13.3* (11.4 to 15.2) (167) |
| Alcohol-related disease | 12.4 (6.7 to 18.1) (16) | 16.3 (5.2 to 27.3) (7) | 15.0 (−0.6 to 30.6) (3) | – | 12.9* (8.3 to 17.5) (26) |
| Bedridden | 13.8* (8.2 to 19.4) (20) | 14.2* (11.8 to 16.7) (114) | 14.3 (−11.6 to 40.2) (1) | 8.3 (−0.7 to 17.4) (3) | 14.0* (11.8 to 16.1) (138) |
| Diabetes | 8.6 (6.4 to 10.8) (56) | 13.1 (10.8 to 15.4) (108) | 3.9 (1.7 to 6.1) (12) | 7.9 (4.4 to 11.4) (18) | 9.6* (8.4 to 10.9) (194) |
| Dialysis or end-stage KD | 12.3 (6.7 to 18.0) (16) | 17.6 (10.2 to 25.0) (18) | 5.1 (0.2 to 9.9) (4) | 22.2* (8.6 to 35.8) (8) | 13.3* (9.7 to 16.8) (46) |
| Transplant | 22.4* (11.7 to 33.1) (13) | 20.0 (2.5 to 37.5) (4) | 3.8 (0.1 to 7.4) (4) | 3.7 (−3.4 to 10.8) (1) | 10.4 (6.3 to 14.6) (22) |
| Epilepsy | 11.3 (6.2 to 16.3) (17) | 14.4 (8.2 to 20.6) (18) | 22.2 (−4.9 to 49.4) (2) | – | 12.9* (9.0 to 16.8) (37) |
| Obesity | 4.0 (0.6 to 7.5) (5) | 15.3 (7.0 to 23.6) (11) | 2.5 (0.1 to 4.9) (4) | 7.1 (−2.4 to 16.7) (2) | 5.7 (3.4 to 8.1) (22) |
| Transfusion | 13.0 (7.6 to 18.5) (19) | 15.0* (11.7 to 18.3) (68) | – | – | 14.4* (11.6 to 17.1) (87) |
| Anaemia | 11.7* (8.8 to 14.6) (56) | 13.4 (11.4 to 15.4) (148) | 7.0* (3.7 to 10.3) (16) | 11.3* (8.5 to 14.2) (55) | 12.0* (10.7 to 13.3) (275) |
| COVID | 4.5 (2.9 to 6.1) (29) | 9.4 (7.7 to 11.1) (111) | 2.3 (−2.1 to 6.7) (1) | 8.9 (0.6 to 17.2) (4) | 7.6 (6.4 to 8.8) (145) |
| Stoma | 13.6 (3.5 to 23.8) (6) | 21.3 (9.6 to 33.0) (10) | 9.6* (4.8 to 14.4) (14) | 11.1 (4.3 to 18.0) (9) | 12.3* (8.7 to 15.9) (39) |
| COPD | 16.3* (11.3 to 21.3) (34) | 12.0 (8.9 to 15.1) (51) | 7.1 (−2.4 to 16.7) (2) | 9.4 (−0.7 to 19.5) (3) | 13.0* (10.5 to 15.5) (90) |
| Cancer - primary | 13.2* (10.0 to 16.5) (56) | 11.3 (8.5 to 14.1) (56) | 3.5 (2.5 to 4.4) (51) | 7.0 (5.0 to 9.0) (44) | 6.9 (6.0 to 7.8) (207) |
| Cancer - metastatic | 18.8* (13.1 to 24.5) (34) | 12.0 (7.5 to 16.5) (24) | 9.2* (5.6 to 12.8) (23) | 7.4 (1.7 to 13.1) (6) | 12.2* (9.8 to 14.6) (87) |
| Pneumonia | 6.2 (4.6 to 7.7) (57) | 12.7 (11.2 to 14.2) (238) | 12.8 (2.3 to 23.3) (5) | 14.1 (5.5 to 22.6) (9) | 10.6* (9.5 to 11.8) (309) |
| UTI | 8.7 (5.8 to 11.6) (32) | 13.9* (11.8 to 16.0) (142) | 8.0 (−2.6 to 18.6) (2) | 9.3 (0.6 to 18.0) (4) | 12.4* (10.7 to 14.0) (180) |
| Length of stay - 15+ days | 11.4* (9.6 to 13.2) (137) | 14.2* (12.7 to 15.8) (278) | 8.7* (6.5 to 10.9) (55) | 10.4* (8.2 to 12.5) (80) | 12.1* (11.1 to 13.0) (550) |
| Discharge_day - weekday | 7.6 (6.9 to 8.3) (419) | 11.2 (10.4 to 12.0) (663) | 2.6 (2.2 to 2.9) (171) | 6.5 (5.5 to 7.4) (167) | 6.9 (6.5 to 7.2) (1420) |
| Discharge_day - weekend | 7.8 (6.8 to 8.8) (217) | 11.1 (10.0 to 12.2) (351) | 2.5 (2.1 to 3.0) (118) | 6.3 (5.1 to 7.6) (91) | 6.5 (6.0 to 6.9) (777) |
| Total | 7.7 (7.1 to 8.3) (636) | 11.1 (10.5 to 11.8) (1014) | 2.6 (2.3 to 2.8) (289) |
6.4 (5.7 to 7.2) (258)
|
6.7 (6.4 to 7.0) (2197) |
p < 0.05
COPD, chronic obstructive pulmonary disease; KD, kidney disease; UTI, urinary tract infection.
At the multivariate level (figure 1), after correcting for age and gender, alcohol-related disease had the highest OR for readmission (OR=2.06, CI 1.36 to 3.13), followed by metastatic cancer (OR=1.98, CI 1.57 to 2.50), epilepsy (OR=1.93, CI 1.36 to 2.75), dialysis or end-stage kidney disease (OR=1.92, CI 1.39 to 2.66), COPD (OR=1.88, CI 1.49 to 2.36), stoma (OR=1.72, CI 1.22 to 2.44), transplant (OR=1.62, CI 1.03 to 2.55), being bedridden (OR=1.57, CI 1.28 to 1.93), anaemia (OR=1.57, CI 1.35 to 1.83), UTI (OR=1.54, CI 1.30 to 1.83), pneumonia (OR=1.52, CI 1.31 to 1.75), dementia (OR=1.49, CI 1.24 to 1.79), diabetes (OR=1.37, CI 1.17 to 1.61) and transfusion (OR=1.34, CI 1.03 to 1.73). COVID, when adjusted for other variables, appears to have a protective effect on readmissions (OR=0.80, CI 0.66 to 0.98).
Figure 1. Logistic regression model results (OR and 95% CI). COPD, chronic obstructive pulmonary disease; KD, kidney disease; UTI, urinary tract infection.
Discussion
This retrospective study based on routinely collected administrative data identified several conditions that, in addition to demographic and admission-related characteristics, significantly increased the risk of readmission within 30 days. Particularly, chronic conditions, such as alcohol-related disease, dialysis or end-stage kidney disease, COPD and dementia, acute conditions like pneumonia and UTIs, and medical devices or therapies (including stoma, dialysis, transplant and transfusion) were associated with a higher risk of readmission to acute care. Cancer, epilepsy, being bedridden or anaemia were also linked to readmissions.
The present study found an overall readmission rate of 6.7% (6.4%–7.0%). A similar study on discharge records conducted in Belgium in 200818 found a readmission rate of 5.2%; the ageing of the population that has occurred over the last 15 years could explain the increase in readmission rates found in this study. Other studies focusing on specific groups of patients found rates ranging from 5% to 30%,19,23 according to type of index admission, age, main disease and comorbidities. Most of these findings are not comparable with those of the present study because they analyse specific clinical conditions, with the exception of one study conducted in Singapore that found a higher readmission rate for medical patients (14.7%)22 and another that, like the present study, found no difference in readmissions between weekend and weekday discharges.23
Analysing specific clinical conditions that increase the risk of readmission, our figures indicate that patients with cancer, especially metastatic cancer, have a higher probability of 30-day readmission. Cancer is known to increase the risk of readmission due to complications of both illness and surgery or medical therapy.10 To identify the cause, Zibelli et al10 interviewed patients with a cancer diagnosis during their hospital readmission. Most patients saw their readmission as caused by problems that could not be treated in an outpatient setting. Others felt they did not have sufficient resources at home to manage their care. Furthermore, the patients did not see the outpatient care team as a resource that they could call on for assistance. Our figures indicate that patients with cancer, especially metastatic cancer, have a higher probability of 30-day readmission. This figure includes both patients who actually and appropriately needed acute palliative interventions and patients who could have been managed in an outpatient, hospice or home care setting. Improving the culture of palliative care both in the hospital setting (with a special focus on communicating diagnosis and prognosis to the patient and the family), in the nursing home setting, and among general practitioners (where some resistance, fear or inexperience in the use of pain management drugs and other palliative care measures may exist) is crucial.24
The present study revealed that diseases with significant social connotations, such as alcohol-related diseases and COPD - stemming from substance abuse and often linked to social deprivation - exhibit higher rates of readmission. In comparison, Morton et al25 found that nearly one-third of patients with COPD are readmitted within 28 days of discharge. COPD exacerbations frequently lead to new hospitalisations, often due to infections and/or respiratory insufficiency. Skov et al26 reported that alcohol-related diseases increased both emergency department visits and length of stay, particularly for those living alone. This is an everyday challenge for clinicians and public health doctors, as hospital readmissions for social reasons are common, leading to unnecessary prolonged hospital stays and complicating discharge planning, which often results in institutionalisation at a young age. Policies should tackle acute social issues, such as the unforeseen loss of a caregiver, with at least temporary, immediate institutionalisation in local facilities. Enhancing services to mitigate substance abuse could also help reduce the escalation of substance-related illnesses, thereby decreasing the necessity for hospitalisation.
Diabetes is also linked to a higher risk of readmission in our sample. Rubin et al27 developed and validated a tool to predict hospital readmission risk for patients with diabetes. In fact, diabetes is a complex disease that involves both social and health components, resulting in numerous complications that may lead to disability, chronic kidney disease, cardiovascular issues and an increased risk of infections.
Several studies have identified low haemoglobin levels as a predictor of readmission.27 28 Our findings corroborate this association in relation to anaemia and transfusions recorded during the initial admission. From a policy standpoint, it may be advisable that patients with anaemia be subject to closer postdischarge follow-up, and that transfusions administered in outpatient or long-term care (LTC) settings be made more readily accessible. Nevertheless, the establishment of protocols for patient identification and the requisite medical oversight during transfusion procedures pose significant challenges to their effective implementation.
Our figures also indicate that patients with a stoma have a higher probability of 30-day readmission. Forsmo et al29 focused on patients with stomas by implementing nursing education and rehabilitation programmes to reduce adverse outcomes such as length of hospital stay and readmission rates. Indeed, a personalised rehabilitation plan at discharge has been consistently associated with reduced readmission rates in several pathological conditions.14 15
We found that being bedridden and having dementia were predictors of 30-day readmissions. Recognising non-oncological palliative care approaches for the elderly, both in the acute hospital setting and LTC, would help avoid inappropriate treatments and hospitalisations. Additionally, broader use of advanced care planning would assist clinicians in their decision-making and reduce the number of rehospitalisations. Enhancing social care services provided by municipalities, along with appropriate referrals to LTC (even for temporary periods if suitable), can help reduce readmission rates.
Study limitations
This study has several methodological limitations. First, the retrospective cohort design based on administrative hospital discharge records is inherently subject to coding errors and misclassification, which may affect the accuracy of diagnoses, procedures and comorbidities identified using ICD-9 codes. Second, the use of routinely collected data limits the availability of clinical detail, preventing the assessment of disease severity, functional status and other relevant factors that may influence readmission risk. Third, information on socioeconomic characteristics, outpatient care and postdischarge support was not available, potentially resulting in residual confounding. Fourth, readmissions occurring outside the hospitals included in the administrative database may not have been captured, leading to possible underestimation of the true readmission rate. Finally, although the statistical models accounted for clustering at the hospital level, causal inferences cannot be drawn due to the observational nature of the study.
Conclusions
The issue of readmissions to hospitals must be accompanied by a thorough understanding of all the factors that drive them, from social influences to healthcare-related dynamics. We risk missing the broader context if we approach the problem solely from a healthcare perspective. Furthermore, we must recognise that enhancing care may lead to increased readmissions, as it can result in fewer deaths and a higher number of sequelae.30 While some studies suggest that sustained efforts can slightly lower readmission rates, others indicate that interventions aimed at improving care coordination and facilitating access to follow-up care can increase readmission rates, likely because of better access to necessary care.31 These interventions should not be regarded as failures.9 Finally, strengthening the relationship between the care team and the patient to ensure care is provided in the most appropriate setting could help reduce readmissions. Potential interventions also include enhancing transitional care through improved discharge planning, incentivising general practitioners to take on the role of case managers, and using telehealth services (eg, telemonitoring, telephone support) to connect patients with specialist care.10
Footnotes
Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Prepublication history for this paper is available online. To view these files, please visit the journal online (https://doi.org/10.1136/bmjopen-2025-103595).
Data availability free text: Data are available upon reasonable request to corresponding author.
Patient consent for publication: Not applicable.
Provenance and peer review: Not commissioned; externally peer reviewed.
Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Ethics approval: The dataset was pseudo-anonymised before analysis. The ULSS 2 Marca Trevigiana Ethics Committee approved the study protocol (Prot. N. 0022516/24; 07/11/2024). The ethics committee waived the requirement for informed consent to participate.
Data availability statement
Data are available upon reasonable request.
References
- 1.Ibrahim AM, Dimick JB. A Decade Later, Lessons Learned From the Hospital Readmissions Reduction Program. JAMA Netw Open. 2019;2:e194594. doi: 10.1001/jamanetworkopen.2019.4594. [DOI] [PubMed] [Google Scholar]
- 2.Pandey A, Golwala H, Hall HM, et al. Association of US Centers for Medicare and Medicaid Services Hospital 30-Day Risk-Standardized Readmission Metric With Care Quality and Outcomes After Acute Myocardial Infarction. JAMA Cardiol. 2017;2:723. doi: 10.1001/jamacardio.2017.1143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Dang T, Chan W, Khawaja S, et al. Hospital costs for unplanned re‐admissions within 30 days of hospitalisations with heart failure, Australia, 2013–2017: a retrospective cohort study. Medical Journal of Australia. 2024;221:317–23. doi: 10.5694/mja2.52424. [DOI] [PubMed] [Google Scholar]
- 4.Berenson RA, Paulus RA, Kalman NS. Medicare’s Readmissions-Reduction Program — A Positive Alternative. N Engl J Med. 2012;366:1364–6. doi: 10.1056/NEJMp1201268. [DOI] [PubMed] [Google Scholar]
- 5.Boccuti C, Casillas G. Aiming for fewer hospital u-turns: the medicare hospital readmission reduction program. 2017
- 6.Sabbatini AK, Joynt-Maddox KE, Liao JM, et al. Accounting for the Growth of Observation Stays in the Assessment of Medicare’s Hospital Readmissions Reduction Program. JAMA Netw Open. 2022;5:e2242587. doi: 10.1001/jamanetworkopen.2022.42587. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Figueroa JF, Wadhera RK. A Decade of Observing the Hospital Readmission Reductions Program—Time to Retire an Ineffective Policy. JAMA Netw Open. 2022;5:e2242593. doi: 10.1001/jamanetworkopen.2022.42593. [DOI] [PubMed] [Google Scholar]
- 8.Zuckerman RB, Sheingold SH, Orav EJ, et al. Readmissions, Observation, and the Hospital Readmissions Reduction Program. N Engl J Med. 2016;374:1543–51. doi: 10.1056/NEJMsa1513024. [DOI] [PubMed] [Google Scholar]
- 9.Joynt KE, Jha AK. Thirty-Day Readmissions — Truth and Consequences. N Engl J Med . 2012;366:1366–9. doi: 10.1056/NEJMp1201598. [DOI] [PubMed] [Google Scholar]
- 10.Zibelli A, Holland K, Wei E. Causes of Cancer Re-Admissions: A Patient-Centered Approach. JCO Oncology Practice . 2020;16:e734–40. doi: 10.1200/JOP.19.00518. [DOI] [PubMed] [Google Scholar]
- 11.Green H, Yahav D, Eliakim-Raz N, et al. Risk-factors for re-admission and outcome of patients hospitalized with confirmed COVID-19. Sci Rep. 2021;11:17416. doi: 10.1038/s41598-021-96716-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.McLaren RA, Magenta M, Gilroy L, et al. Predictors of readmission for postpartum preeclampsia. Hypertens Pregnancy. 2021;40:254–60. doi: 10.1080/10641955.2021.1975737. [DOI] [PubMed] [Google Scholar]
- 13.McIntyre LK, Arbabi S, Robinson EF, et al. Analysis of Risk Factors for Patient Readmission 30 Days Following Discharge From General Surgery. JAMA Surg. 2016;151:855–61. doi: 10.1001/jamasurg.2016.1258. [DOI] [PubMed] [Google Scholar]
- 14.Chen Y-W, Wang C-Y, Lai Y-H, et al. Home-based cardiac rehabilitation improves quality of life, aerobic capacity, and readmission rates in patients with chronic heart failure. Medicine (Baltimore) 2018;97:e9629. doi: 10.1097/MD.0000000000009629. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kumar A, Resnik L, Karmarkar A, et al. Use of Hospital-Based Rehabilitation Services and Hospital Readmission Following Ischemic Stroke in the United States. Arch Phys Med Rehabil. 2019;100:1218–25. doi: 10.1016/j.apmr.2018.12.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kansagara D, Englander H, Salanitro A, et al. Risk prediction models for hospital readmission: a systematic review. JAMA. 2011;306:1688–98. doi: 10.1001/jama.2011.1515. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Zhou H, Della PR, Roberts P, et al. Utility of models to predict 28-day or 30-day unplanned hospital readmissions: an updated systematic review. BMJ Open. 2016;6:e011060. doi: 10.1136/bmjopen-2016-011060. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Braet A, Weltens C, Sermeus W, et al. Risk factors for unplanned hospital re‐admissions: a secondary data analysis of hospital discharge summaries. Evaluation Clinical Practice. 2015;21:560–6. doi: 10.1111/jep.12320. [DOI] [PubMed] [Google Scholar]
- 19.Albinali HAH, Singh R, Al Arabi AR, et al. Predictors of 30-Day Re-admission in Cardiac Patients at Heart Hospital, Qatar. Heart Views. 2023;24:125–35. doi: 10.4103/heartviews.heartviews_91_22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Brooke BS, Stone DH, Cronenwett JL, et al. Early Primary Care Provider Follow-up and Readmission After High-Risk Surgery. JAMA Surg. 2014;149:821. doi: 10.1001/jamasurg.2014.157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Buldu I, Tepeler A, Karatag T, et al. Which factors affect the hospital re-admission and re-hospitalization after flexible ureterorenoscopy for kidney stone? World J Urol. 2016;34:1291–5. doi: 10.1007/s00345-015-1750-x. [DOI] [PubMed] [Google Scholar]
- 22.Chan MF, Wong FKY. The risk factors for hospital re-admission in medical patients in Singapore. Health Soc Care Community . 2014;22:488–96. doi: 10.1111/hsc.12108. [DOI] [PubMed] [Google Scholar]
- 23.Chiou L-J, Chen H-M, Pan L-F, et al. Holiday ratio of hospitalization and 30‐day readmission rates among cancer patients after major surgery. Cancer Med. 2022;11:743–52. doi: 10.1002/cam4.4482. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Al Sattouf A, Farahat R, Khatri AA. Effectiveness of Transitional Care Interventions for Heart Failure Patients: A Systematic Review With Meta-Analysis. Cureus. 2022;14:e29726. doi: 10.7759/cureus.29726. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Morton K, Sanderson E, Dixon P, et al. Care bundles to reduce re-admissions for patients with chronic obstructive pulmonary disease: a mixed-methods study. Health Serv Deliv Res. 2019;7:1–128. doi: 10.3310/hsdr07210. [DOI] [PubMed] [Google Scholar]
- 26.Skov NF, Tygesen GB, Lisby M. Associations Between Readmission and Length of Stay in the Acute Admission Unit for Patients With Alcohol-Related Diagnoses: A Cohort Study. J Stud Alcohol Drugs. 2025;86:402–9. doi: 10.15288/jsad.23-00395. [DOI] [PubMed] [Google Scholar]
- 27.Rubin DJ, Handorf EA, Golden SH, et al. Development and Validation of a Novel Tool to Predict Hospital Readmission Risk Among Patients with Diabetes. Endocr Pract. 2016;22:1204–15. doi: 10.4158/E161391.OR. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Melvin JC, Smith JB, Kruse RL, et al. Risk Factors for 30-Day Hospital Re-Admission with an Infectious Complication after Lower-Extremity Vascular Procedures. Surg Infect (Larchmt) 2017;18:319–26. doi: 10.1089/sur.2016.234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Forsmo HM, Pfeffer F, Rasdal A, et al. Pre- and postoperative stoma education and guidance within an enhanced recovery after surgery (ERAS) programme reduces length of hospital stay in colorectal surgery. Int J Surg. 2016;36:121–6. doi: 10.1016/j.ijsu.2016.10.031. [DOI] [PubMed] [Google Scholar]
- 30.Gorodeski EZ, Starling RC, Blackstone EH. Are all readmissions bad readmissions? N Engl J Med. 2010;363:297–8. doi: 10.1056/NEJMc1001882. [DOI] [PubMed] [Google Scholar]
- 31.Weinberger M, Oddone EZ, Henderson WG. Does increased access to primary care reduce hospital re-admissions? Veterans Affairs Cooperative Study Group on Primary Care and Hospital Re-admission. N Engl J Med. 1996;334:1441–7. doi: 10.1056/NEJM199605303342206. [DOI] [PubMed] [Google Scholar]