Avoiding unplanned readmissions following discharge from inpatient surgery is a priority for hospitals, particularly when treating patients over the age of 65.
Patients in this age group undergo surgical procedures more frequently and have a higher rate of readmission1 . Hospitals are motivated to prevent readmission among Medicare patients receiving hip or knee arthroplasty, as well as those receiving coronary artery bypass graft (CABG) surgery, by the Hospital Readmissions Reduction Program (HRRP).2 Unplanned readmission within 30 days of a surgical procedure can disrupt care plans and complicate rehabilitation for these patients. Sepsis is a major risk factor for readmissions among elder patients recovering from surgery, with age and the presence of open wounds known to be contributing factors.3
Medicare beneficiaries requiring post-acute care are most often discharged to either a skilled nursing facility (SNF) or home, with care from a home health (HH) provider.3 Research examining readmission among these patients has yielded mixed results. At least one large observational study found that discharge to SNF was associated with higher rates of 30-day readmission by 6.6 percentage points,1 while another found that patients discharged to HH were readmitted more frequently by 5.6 percentage points.4
These findings suggest several implications for hospital discharge planning. While both SNF and HH are valuable options for post-acute care, hospitals should carefully consider the specific needs of each patient when making discharge decisions. Patients who require more intensive, round-the-clock care, such as those recovering from orthopedic procedures, may be appropriate for SNF, particularly when mobility may present issues at home. Patients at higher risk for infection, due to individual history or the type of procedure performed, may be more appropriate for HH.
Claims were retrieved from the Washington Discharge Dataset for patients aged 65 and up who had an inpatient procedure between January 1, 2022 and July 31, 2024 in either the operating room (OR) department or a General Surgery service line, with discharge to either SNF or HH. Characteristics of the initial discharge included the primary procedure category, length of stay (LOS), and severity/mortality risk of primary diagnosis using APR-DRG classification. Outcomes were unplanned readmission to acute IP or ED within 30 days of discharge; secondary outcomes were readmissions with sepsis, defined as general sepsis, severe sepsis, or septic shock.
Results
Discharge characteristics are summarized in Table 1 and Figure 1. Patients were more likely to be discharged to SNF (62.2% of discharges) rather than HH (37.8%) following OR procedure and/or General Surgery. Those discharged to SNF had higher geometric mean LOS (8.1 days) compared to those discharged to HH (6.1 days). Although APR DRG Severity and Mortality Risk were similar among these groups, a greater proportion patients discharged to SNF were more likely to have Level 4 Severity and Mortality Risk (19% and 17%, respectively) than those discharged to HH (15% and 14%, respectively). Patients discharged to SNF also tended to be older, with median age 78 compared to 75 for HH.
Procedures
Table 2 shows the procedures most frequently performed in the inpatient visit preceding discharge to either HH or SNF. In addition to the count of each procedure, the proportion of patients discharged to either setting who received that procedure is shown. Femur fixation and hip arthroplasty were the most common procedures preceding discharge to SNF, together accounting for more than 27% of all discharges to SNF. While these procedures were also the most common among patients discharged to HH, together they account for only about 10% of discharges.
Figure 2 shows the proportion of patients discharged to SNF or HH following each of the most common procedures. Patients receiving skeletal procedures like femur/leg/foot fixation or hip arthroplasty were discharged to SNF more than 75% of the time. Patients receiving coronary artery bypass grafts (CABG) or colectomy were more likely to be discharged to HH.
Readmissions
Tables 3a and 3b summarize the rates of unplanned readmission to acute IP or ED visit, respectively, within 30 days of discharge to either SNF or HH. Additionally, rates of sepsis are shown as a percentage of IP readmissions or visits to ED.
Among patients in this analysis, rates of acute IP readmission were nearly identical, though slightly higher among patients discharged to SNF (15%) compared to HH (14.8%). And among patients who were readmitted, readmission with sepsis was more common among patients readmitted after discharge to SNF (21.6% of readmissions) rather than HH (14.7% of readmissions). Conversely, HH patients had ED visits within 30 days of discharge at a higher rate (25.9%) than those discharged to SNF (23.2%).
Figure 3 shows how rates of IP readmission vary by procedure type between discharge settings, ordered by total volume of the procedure. SNF discharges are more likely to result in readmission for all but one procedure: urinary tract dilation. The difference in rates of readmission is noticeably higher for spine fusion, knee arthroplasty, and CABG.
Discussion
This analysis of 74,797 discharges to either SNF or HH among older surgical patients revealed several key findings. Notably, readmission rates were comparable between these two post-acute care settings. However, SNF-discharged patients appeared to be more likely to experience sepsis during readmission.
While patients in this cohort were more often discharged to SNF than HH, the type of procedure performed appeared to influence the choice of discharge setting. Procedures affecting the skeletal system and/or mobility more often preceded discharge to SNF, while internal procedures more commonly preceded HH discharge. Readmission rates were noticeably lower among patients discharged to either setting after receiving hip or knee arthroplasty, which are included in the Hospital Readmissions Reduction Program (HRRP) for Medicare patients.
Existing studies have shown differential rates of readmission between patients discharged to SNF versus HH, but readmission rates in this cohort were virtually identical between discharge settings. Comparing rates of discharge to HH or SNF among procedure groups seems to suggest that procedures affecting mobility are more likely to result in discharge to SNF, suggesting a greater need for round-the-clock care among these patients. At the same time, higher rates of sepsis among patients readmitted from SNF may reflect a greater risk of infection when treated in a facility around other patients, particularly for those with surgical wounds and/or those recovering from internal procedures.
Conclusions
These findings suggest several implications for hospital discharge planning. While both SNF and HH are valuable options for post-acute care, hospitals should carefully consider the specific needs of each patient when making discharge decisions. Patients who require more intensive, round-the-clock care, such as those recovering from orthopedic procedures, may be appropriate for SNF, particularly when mobility may present issues at home. Patients at higher risk for infection, due to individual history or the type of procedure performed, may be more appropriate for HH.
Hospitals may benefit from analyzing rates of readmission rates of their own discharges to better understand trends in their patient population. To avoid sepsis readmissions, hospitals should prioritize infection prevention and control strategies, particularly for patients discharged to SNF. Hospitals could work with SNFs to develop early warning systems for sepsis, allowing for prompt identification and treatment, potentially avoiding acute inpatient admission.
By carefully considering these factors, hospitals can improve discharge planning and reduce readmission rates, ultimately improving patient outcomes and reducing healthcare costs.
Limitations
It is important to note that this study has several limitations. Firstly, this analysis is exploratory in nature and lacks formal statistical testing. Thus, the observed relationships between discharge disposition and readmission rates are sensitive to random variation. Because of this, findings should be interpreted cautiously and require further validation through more rigorous statistical analyses.
Secondly, this analysis did not account for a variety of patient factors that may influence readmission risk. While patients in this study cohort are quite similar in terms of age, APR DRG severity, and mortality risk, there are likely to be important differences between these patient populations. SNF-discharged patients tended to be older than those discharged to HH. Additionally, SNF-discharged patients had longer mean LOS, possibly indicating greater complexity. The decision to discharge to either of these settings may be influenced by Medicare reimbursement, individual complications, as well as input from the patient and their families. Further, the decision to discharge to HH rather than SNF may be influenced by geographic proximity to care facilities. Future analyses of this topic may benefit from examining the roles of insurance payor and patient or hospital rurality in rates of discharge to and readmission from SNF and HH.
2 Medicare.gov. Hospital Readmissions Reduction Program (HRRP). Accessed November 18, 2024. https://www.cms.gov/medicare/payment/prospective-payment-systems/acute-inpatient-pps/hospital-readmissions-reduction-program-hrrp.
3 Flanagan JM, et al. Factors associated with the rate of sepsis after surgery. Crit Care Nurse 2020;40(5):e1-e9. doi:10.4037/cnn2020171
4 Werner RM, et al. Patient outcomes after hospital discharge to home with home health care vs to a skilled nursing facility. JAMA Intern Med 2019;179(5):617:623. doi:10.1001/jamainternmed.2018.7998
5Medicare.gov. Home health services. Accessed November 18, 2024. https://www.medicare.gov/coverage/home-health-services
