Unplanned readmissions are a major concern for hospitals. There are myriad reasons why a patient might return to the hospital unexpectedly, ranging from substandard care, poor discharge planning, ineffective post-discharge services and more1
Social determinants of health like demographics, income, education and English proficiency have a significant impact on readmission rates as well, especially among Medicaid patients.2 Due to the complexity of unplanned readmissions, we will limit this analysis to secondary diagnoses during unplanned readmissions.
A primary diagnosis is typically the most severe or demanding diagnosis, while secondary diagnoses can be any coexisting condition during the same stay. The same secondary diagnoses are present alongside many different primary diagnoses, indicating a potential relationship between secondary diagnoses and unplanned readmissions. This relationship can be viewed from the initial stay to the readmission and in reverse. For example, investigating correlations between common secondary diagnoses and initial visit length of stay.
This analysis uses claims data from the Washington Discharge Dataset for all patients who had an inpatient stay between Jan. 1, 2022 and Sept. 30, 2024 with an unplanned readmission within 30 days after discharge. This population includes over 130,000 individual claims.
In this dataset, secondary diagnoses are not listed in any particular order, though we assume more severe secondary diagnoses will be listed first, and use only the first ten diagnoses listed on a claim for this analysis. Diagnoses are grouped according to the Clinical Classifications Software (CCS) for ICD-10 diagnosis coding.3 Unplanned readmissions involving newborns and rehabilitation discharges are excluded for the purposes of this analysis.
Analysis
A large portion of patients with unplanned readmissions have the same primary diagnoses on their initial discharge. The ten most frequent initial primary diagnosis groups are summarized in Table 1. These account for just under 45% of all claims that result in an unplanned readmission.
Of these patients, we then examine the most common secondary diagnoses from their readmission encounter. The ten most frequent secondary diagnosis readmission groups are given in Table 2. Note that the total percentage of all diagnosis groups can be greater than 100%, as an individual claim can have multiple secondary diagnoses.
From this, we can see congestive heart failure, fluid and electrolyte disorders, and hypertension with complications and secondary hypertension each appear on about a third of these claims. Figures 1, 2 and 3 show how these secondary readmission diagnosis groups are distributed among our set of most frequent initial primary diagnosis groups. Hypertension with complications and septicemia are standouts in all three cases.
We can also look at the length of the initial stay which resulted in an unplanned readmission by whether the readmission included our subset of secondary diagnoses or not. Figure 4 shows the average length of stay days for initial stay diagnoses grouped by whether they include our most significant secondary diagnoses (In) or not (Out) in the resultant readmission. Except for respiratory failure, all readmissions with these secondary diagnoses had increased average length of stay.
Conclusion
By better understanding the markers of unplanned readmissions, hospitals can identify higher-risk secondary diagnoses which may lead to an unplanned readmission.
Likewise, an increased focus on treating the most common secondary diagnoses such as congestive heart failure, hypertension with complications, and fluid and electrolyte disorders in the initial encounter could help reduce the frequency of unplanned readmissions.
Further exploration into specific primary diagnoses and the secondary diagnoses that frequently accompany them could yield better insights into disease profiles that are more effectively treated together.
____________________
1CMS Office of Minority Health (2024, April). Guide for Reducing Disparities in Readmissions. https://www.cms.gov/about-cms/agency-information/omh/downloads/omh_readmissions_guide.pdf.
2Agency for Healthcare Research and Quality (2016, September). Designing and Delivering Whole-Person Transitional Care: The Hospital Guide to Reducing Medicaid Readmissions. https://www.ahrq.gov/sites/default/files/wysiwyg/professionals/systems/hospital/medicaidreadmitguide/medicaidreadmissions.pdf.
3Hcup-us.ahrq.gov. Healthcare Cost & Utilization Project User Support. Accessed February 12, 2025. https://hcup-us.ahrq.gov/datainnovations/icd10_resources.jsp.
