IMPACT OF INPATIENT PULMONARY MEDICINE CONSULTATION FOR COPD EXACERBATION ON 90-DAYS AND 1-YEAR HOSPITAL READMISSION RATES

Abstract

TOPIC: Obstructive Lung Diseases TYPE: Original Investigations PURPOSE: In 2010, the annual national medical costs from COPD were estimated at $32 billion. COPD is the third leading cause of death in the U.S. and exacerbations carry a poor prognosis with a 1-yr mortality of 22%. We investigated if inpatient pulmonary intervention reduced readmission rates. METHODS: Adults admitted with a diagnosis of COPD between Nov 2015 and Dec 2018 were identified using Clinical Looking Glass, institutional software for data collection. 900 patients were screened and those who had no PFTs in the system or were deemed not to be admitted for COPD were excluded from analysis. 330 patients met inclusion criteria, 124 patients were seen by pulmonary medicine, 206 patients had no pulmonary consultation during admission. Median length of stay (LOS) and interquartile range (IQR) were generated and distributions across groups were compared via Wilcoxon Mann-Whitney tests. 90-day and 1-yr readmission rates were compared between groups via chi-square test. Two-sided p-values <0.05 were statistically significant. Logistic regression models were estimated to explore factors associated with pulmonary consultation. Odds ratios for the probability of having a pulmonary consultation were generated along with 95% confidence intervals. RESULTS: There were 82 (25%) 90-day readmissions; 45 (22%) in no consultation group, 37 (30%) in consultation group (p=0.12). There were 152 (46%) 1-yr readmissions; 90 (44%) in no consultation group, 62 (51%) in consultation group (p=0.31). Median LOS overall was 4 days (IQR 3-7); LOS in the no consultation group was shorter than in the consultation group 4 days (3-5) vs 6 (4-8) with p<0.0001. In a joint model, older patients (OR 1.03, 95% CI 1.004, 1.050; p=0.02) and those with lower FEV1% (OR 1.01; 95% CI 1.003, 1.026; p=0.01) had higher odds of consultation. We also analyzed LOS and readmission rates for patients on home oxygen (O2) or home noninvasive ventilation (NIV). Of the 328 with available data, 76 (23%) were on O2 and 252 (77%) were not. Median LOS (IQR) for patients on O2 was 5 days (3-7) compared to 4 days (3-6) for patients not on O2 (p=0.16). A total of 30 patients (40%) in O2 group had readmission within 90 days compared to 51 (20%) in no O2 group (p=0.001); 1-yr readmission rates were also associated with O2 status with 46 (61%) in O2 group, 104 (41%) in no O2 group (p=0.004). Of the 327 with available data, 72 (22%) patients were on NIV at home. Median LOS (IQR) for patients on NIV was 5 days (4-8) compared to 4 days (3-6) for no NIV (p=0.002). The NIV group had a higher 90-day readmission rate but this was not statistically significant (NIV 31% vs no NIV 23%; p=0.22). Similarly, there was an increase in 1yr readmission rates in the NIV group (NIV 55% vs no NIV 43%) but this was not statistically significant (p=0.08). CONCLUSIONS: Pulmonary consult remains a statistically insignificant predictor of 90-day and 1-yr readmission. Home 02 was associated with higher 90-days and 1-yr readmission rates, LOS was statistically non-significant. Being on home NIV was associated with a longer hospital LOS but 90-days and 1-yr readmission rates were not statistically significant. CLINICAL IMPLICATIONS: Although in our study pulmonary consult remains a statistically insignificant predictor for readmission, the consultation was only sought for patients with worse baseline disease. Pulmonary intervention early in the disease may be beneficial for disease management and prevention of exacerbations. DISCLOSURES: No relevant relationships by SYED ABBAS, source=Web Response No relevant relationships by Krystal Cleven, source=Web Response No relevant relationships by Aditi Desai, source=Web Response No relevant relationships by Melissa Fazzari, source=Web Response No relevant relationships by Jonathan LaMee, source=Web Response No relevant relationships by Gordon Pelegrin, source=Web Response No relevant relationships by Rizwan Zafar, source=Web Response