Author ORCID Identifier

https://orcid.org/0000-0003-0656-3069

Semester

Summer

Date of Graduation

2023

Document Type

Dissertation

Degree Type

PhD

College

School of Pharmacy

Department

Pharmaceutical Systems and Policy

Committee Chair

Khalid M. Kamal, MPharm., PhD, Chair

Committee Member

George A. Kelley, DA, FACSM

Committee Member

SiJin Wen, PhD, MS, MA

Committee Member

Rahul Sangani, MD

Committee Member

Toni Marie Rudisill, PhD

Committee Member

Traci J. LeMasters, PhD

Committee Member

Virginia G. Scott, PhD

Abstract

Background: Predicting the survival of patients with Idiopathic pulmonary fibrosis (IPF) remains challenging due to the remarkable heterogeneity of the disease prognosis. Therefore, developing reliable prognostic tools is of utmost importance in IPF management. Nomograms currently represent the most accurate and discriminatory tools for predicting outcomes, with superior performance compared to the commonly used integer score approach. Therefore, developing a valid local nomogram to predict mortality among patients with IPF is warranted. Additionally, IPF is characterized by a high symptom burden significantly impacting Health-related quality of life (HRQoL). Accurately measuring HRQoL in IPF is crucial as it guides treatment decisions and monitors disease progression, ultimately leading to improved care and outcomes. Despite being widely utilized as instruments to assess HRQoL in IPF, the psychometric properties of the St George's Respiratory Questionnaire (SGRQ) and its IPF version (SGRQ-I) have not been thoroughly evaluated among patients with IPF. Therefore, this dissertation pursued three objectives: 1) to develop a nomogram and nomogram-based scoring system predicting overall survival (OS); 2) to validate a newly developed nomogram-based scoring system against the widely used gender-age-physiology (GAP) model; 3) To conduct a systematic review with meta-analysis to assess and summarize the available evidence on the psychometric properties of the SGRQ and SGRQ-I in IPF patients using the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) framework. Methods: We conducted a retrospective cohort study including 152 patients with IPF identified from the electronic medical records of a West Virginia hospital system. The least absolute shrinkage and selection operator (LASSO) Cox regression analysis was employed for variable selection, and a nomogram was constructed based on the selected independent predictive variables. The nomogram was designed to visually represent the survival prediction model using the variables selected by LASSO, incorporating multiple factors to estimate an individual patient's 5-year survival probability. Risk groups were defined based on the nomogram's probability tertiles. The performance of the nomogram-based model was evaluated using Harrell's concordance index (C-index) and the Hosmer-Lemeshow test, assessing discrimination and calibration, respectively. In 113 patients with IPF, the nomogram-based scoring system's performance was compared to the GAP staging system in terms of discrimination using Harrell's C discrimination index (C-statistic) and Royston-Sauerbrei's D statistics, as well as calibration using the Hosmer-Lemeshow goodness-of-fit test and the Brier score. For the systematic review, a comprehensive search was performed to identify relevant studies from seven databases: PubMed, Medline (via OVID), CINAHL, PsycINFO, Web of Science, Scopus, and the ProQuest Dissertations, up until June 2022. The risk of bias assessment was conducted using the COSMIN risk of bias checklist. The instrument's measurement properties were evaluated based on COSMIN criteria. The quantitative analysis involved effect size calculation and pooling using the Fisher r to z transformation and inverse-variance heterogeneity (IVhet) model, respectively. Heterogeneity and small-study effects were examined, and qualitative summarization was also performed. The strength of evidence was assessed using a modified Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach.

Results: The study population primarily consisted of elderly male patients with a body mass index (BMI) above 24 Kg/m2, reflecting common demographic characteristics in IPF. LASSO regression analysis identified Diffusing capacity of the lungs for carbon monoxide percentage predicted (DLco%), BMI, pulmonary hypertension, pulmonary embolism, and sleep apnea as independent predictive variables for OS in IPF. These variables were included in the nomogram, which demonstrated good discrimination with a C-index of 0.71, indicating its ability to differentiate between patients with different survival outcomes. The nomogram also exhibited good calibration, indicating its accurate estimation of survival probabilities. The nomogram-based scoring system also exhibited similar discriminative properties to the GAP staging system but showed better separation between risk groups, particularly in earlier periods. A sensitivity analysis stratified by the presence of emphysema showed improved separation in the nomogram-based scoring system among IPF patients. As for the systematic review, A total of 24 studies were included in the systematic review, with 19 assessing the psychometric properties of the SGRQ and seven evaluating the SGRQ-I. The content validity of both questionnaires was not adequately assessed in patients with IPF. Structural validity was only evaluated in one study for the SGRQ-I, and insufficient information was available for conclusive judgment. Internal consistency showed acceptable values for most domains of the SGRQ and SGRQ-I, except for the symptom domain of the SGRQ, which had low internal consistency. Test-retest reliability was found to be adequate for both questionnaires. Construct validity showed moderate to strong correlations with respiratory-specific measures but weak correlations with pulmonary function tests. Known-group validity demonstrated significant differences in disease severity measures, supporting the instruments' ability to distinguish between different groups of patients. Both questionnaires demonstrated adequate responsiveness. Minor flooring and ceiling effects were observed. Conclusion: The nomogram-based staging system developed showed promising performance in predicting overall survival and represents an actionable tool that could potentially improve clinical management in IPF. Furthermore, the study emphasizes the significance of routine screening for malnutrition and the presence of pulmonary comorbidities in IPF care management to optimize patient outcomes. The SGRQ and SGRQ-I have the potential to capture important aspects of HRQoL in the IPF population. However, the content validity of both questionnaires has not been assessed in patients with IPF, and the structural validity of the SGRQ and SGRQ-I also requires more comprehensive evaluation. While internal consistency, test-retest reliability, and known-group validity were generally satisfactory, construct validity showed weak correlations with pulmonary function tests. Future studies should aim to establish the content and structural validity of the SGRQ and SGRQ-I in IPF. Such research endeavors will enhance the quality and applicability of these patient-reported outcome measures in clinical practice and research.

Embargo Reason

Publication Pending

Download

Share

COinS