Markers for disease severity : pharmacoepidemiological studies on obstructive lung disease

Abstract

Disease severity is important for treatment of the individual patient, but also in the conduct of pharmacoepidemiological studies. The objective of this thesis was to identify and evaluate objective markers for disease severity with focus on medication use and biomarkers. In this thesis, obstructive lung disease was used as a tool to study markers for disease severity. Markers for disease severity based on medication use can be derived from the PHARMO Record Linkage System, including complete medication histories of more than two million residents in the Netherlands from 1985 onwards. It was found that patients with an exacerbation out of the hospital treated with glucocorticoids only or in combination with antibiotics were at decreased risk of a first hospitalisation and readmission. Usage of glucocorticoids, antibiotics and other respiratory drugs increased prior to hospitalisation. These results could be indicative of the development and/or treatment of an exacerbation. Therefore, using only hospitalisation as marker for disease severity leads to underestimation of the exacerbation rate. There is a need for markers to detect exacerbations in an early phase in order to start treatment as early as possible and possibly prevent hospitalisations for obstructive lung disease. Biomarkers for disease severity were studied using the Utrecht Patient Oriented Database (UPOD), which comprises clinical data of patients of the University Medical Centre Utrecht. Our results suggest that the absolute neutrophil and eosinophil count can be used as a biomarker for disease severity in obstructive lung disease. The results in this study were adjusted for glucocorticoid use. Including all hospitalised patients, irrespective of diagnosis also showed that the observed neutrophilia in systemic glucocorticoid users in clinical practice is probably associated with the underlying disease, rather than glucocorticoid use itself. Moreover, changes in neutrophil morphology can be used as a biomarker for disease severity in obstructive lung disease. Using UPOD, biomarker studies could be conducted on a larger scale, because the morphology parameters are measured and stored automatically. Testing bias should be taken into account in the conduct of biomarker studies using data from routine clinical databases as it was found that testing for disease severity was conducted for more severely ill patients. Requests for the absolute neutrophil count in clinical practice are associated with underlying disease and this could bias the study results. Distribution of diagnostic subgroups and testing guidelines might vary between hospitals, but testing bias is an issue in all hospitals and should be evaluated to be able to adjust for this bias. From the findings presented in this thesis it can be concluded that identification and evaluation of markers for disease severity is essential from a clinical point of view as well as to be able to adjust for confounding by disease severity in epidemiological studies. Replicate studies should be conducted in a prospective, blinded fashion and the accuracy of biomarkers for disease severity should be confirmed. Studies that combine (molecular) clinical, laboratory medicine and pharmacoepidemiological techniques add promising diagnostic biomarkers to the multidisciplinary health care needed for patients with severe obstructive lung disease.