S66 Is hospital readmission an appropriate outcome measure for community-acquired pneumonia?

Abstract

Introduction and Objectives: There is no ideal outcome measure for studies in community-acquired pneumonia (CAP). Readmission to hospital within 30 days is one proposed measure which coincidentally the NHS now imposes financial penalties on Trusts for. We sought to investigate the frequency and validity of this measure for CAP and to examine factors predictive of readmission in CAP. Methods: All adult cases (ICD10 J10eJ18) admitted in 2010 were identified from Trust Information records. Those readmitted within 30 days of discharge were regarded as cases and reason for readmission was ascertained. The two consecutive admissions after each readmission case were used as controls to identify features predictive of readmission. All were validated as CAP by inspection of radiographs and case records. Results: 562 cases were identified. 93 were excluded. 96 (20%) of the remaining 469 died. 55 (12%) were readmitted. Eight of these were excluded and six case notes were lost leaving 41 cases who were compared with 72 controls who had not been readmitted. Of these 113, mean age was 61 (95% CIs 57e65), 59% were male, 85% had one or more comorbid disease, 83% were admitted from their own home, 54% were CURB65 0e1, 26% CURB65 2, 23% CURB65 3e5. Readmission was considered by the admitting physician to be CAPrelated in only 16 (39%), but even in these CRP was raised at readmission in only 81%. Non-CAP reasons for readmission were varied and were distributed across at least 8 disease areas. Only 13 (32)% of all readmissions were considered to have been preventable at the first admission. Age (OR 0.995; 95%CI 0.959 to 1.033), presence of comorbid disease (2.045; 0.415 to 10.089), Charlson comorbidity index (1.082; 0.785 to 1.491), initial length of stay (1.009; 0.984 to 1.034), CURB65>0 (2.003; 0.551 to 7.282) and initial treatment with tazocin (2.502; 0.804 to 7.784) were significantly related to readmission, but CIs all included unity. Conclusion: The low frequency, lack of relationship to CAP in the majority and lack of preventability at the index admission suggest that readmission within 30 days of discharge is not a valid outcome marker for CAP. Age and markers of biological unfitness predict readmission.