Association Between Postdischarge Pulmonologist Visit and Hospital Readmission

Umur Hatipoğlu,Xiaofeng Wang

DOI: https://doi.org/10.1378/chest.15-0807

IF: 9.6

2015-01-01

Chest

Abstract:To the Editor: We read with interest the article by Gavish et al1Gavish R Levy A Dekel OK Karp E Maimon N The association between hospital readmission and pulmonologist follow-up visits in patients with COPD.Chest. 2015; 148: 375-381Abstract Full Text Full Text PDF PubMed Scopus (53) Google Scholar in this issue of CHEST (see page 375) on the association between pulmonologist follow-up and hospital readmission for patients with COPD. The investigators found that an early follow-up appointment with a pulmonologist was associated with a reduced hospitalization rate because of COPD exacerbation. Their work adds to the growing body of nonrandomized cohort studies that emphasize the importance of timely postdischarge follow-up.2Sharma G Kuo YF Freeman JL Zhang DD Goodwin JS Outpatient follow-up visit and 30-day emergency department visit and readmission in patients hospitalized for chronic obstructive pulmonary disease..Arch Intern Med. 2010; 170: 1664-1670Crossref PubMed Scopus (191) Google Scholar, 3Hernandez AF Greiner MA Fonarow GC et al.Relationship between early physician follow-up and 30-day readmission among Medicare beneficiaries hospitalized for heart failure.JAMA. 2010; 303: 1716-1722Crossref PubMed Scopus (775) Google Scholar We would like to point out aspects of the study that render its findings less applicable to the paradigm of readmissions in the United States and also raise potential methodologic concerns. The authors have chosen rehospitalization related to COPD exacerbation as an end point. In the United States, the Hospital Readmissions Reduction Program penalizes hospitals for an increased all-cause readmission rate. All-cause readmission was specifically chosen to emphasize that “readmission for any cause is a concern and measuring diagnosis related admissions may increase susceptibility to gaming.”4Lindenauer PK Normand SL Drye EE et al.Development, validation, and results of a measure of 30-day readmission following hospitalization for pneumonia.J Hosp Med. 2011; 6: 142-150Crossref PubMed Scopus (125) Google Scholar To this end, we were not able to determine how a readmission was adjudicated to be caused by COPD in the study. Second, based on a large study of Medicare claims data, only 27.6% of readmissions after an index COPD admission are because of COPD.5Shah T Churpek MM Coca Perraillon M Konetzka RT Understanding why patients with COPD get readmitted: a large national study to delineate the medicare population for the readmissions penalty expansion.Chest. 2015; 147: 1219-1226Abstract Full Text Full Text PDF PubMed Scopus (137) Google Scholar Therefore, one could argue that a visit with a specialist may not provide the comprehensive intervention needed to prevent most readmissions. Our major methodologic concern regards the variable selection for the two separate multivariate logistic regression models in the article. First, the authors created a multivariate logistic regression model to determine the characteristics of patients who did not attend a follow-up visit. Here, a previous visit with a pulmonologist was associated with a reduction in the risk of not attending a follow-up visit (adjusted OR, 0.82, P = .002). Subsequently, a separate model was built for the examination of 90-day readmission risk factors. This model does not include a previous visit with a pulmonologist as a covariate. We suggest that having had a pulmonologist visit prior to admission, presumably indicating established specialist care, may be important in determining the risk of hospital readmission. We would like to ask the authors to clarify whether a previous visit with a pulmonologist was included as a covariate in the second regression model and in the Cox proportional hazard model. We were also unable to determine whether “higher number of hospitalizations in the previous year” and “lower frequency of office visits in the previous year” were entered into the models as dichotomous variables or count numbers. Dichotomous assessment of these variables requires arguments to justify the cutoffs. ResponseCHESTVol. 148Issue 2PreviewTo the Editor: Full-Text PDF