Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Health Management and Systems Sciences

Degree Program

Public Health Sciences with a specialization in Health Management, PhD

Committee Chair

Little, Bert

Committee Co-Chair (if applicable)

Esterhay, Robert

Committee Member

Carrico, Ruth

Committee Member

Jennings, J'Aime

Author's Keywords

Diabetes; statin; community acquired pneumonia


BACKGROUND: Statins, a class of drugs that treat hyperlipidemia, may have an immunosuppressive effect for patients with community acquired pneumonia (CAP). Retrospective and in vitro studies have suggested an immunomodulatory, antioxidative and anticoagulant effects from statin use in patients with Type 2 Diabetes Mellitus (T2DM) hospitalized for CAP. Prospective studies that have tested any effect of statin therapy on patients with T2DM and CAP have not been found literature. To date, prospective studies showing of the effects statin therapy may have on T2DM patients hospitalized for CAP are not available. METHODS: This dissertation is a secondary analysis using deidentified data collected from the HAPPI Study, a prospective CAP observational study conducted in nine adult acute-care hospitals in Louisville, Kentucky, from 2014-2017. HAPPI patients were grouped by T2DM, prior statin exposure, and age. Decision tree analyses were performed to indicate how strongly the T2DM and statin interaction is related to outcomes of mortality (after one, six, and 12 months) and CAP rehospitalization (after one, six, and 12 months). Multivariate logistic regression analyses were used to identify potential covariables. Propensity score matching (PSM) and the McNemar test were used to compare the odds ratios of outcomes on paired statin users (cases) and non-statin users (controls) based on age and T2DM. RESULTS: From 10052 CAP patients, 1265 of 2734 T2DM patients were on statins (46.3%) and 2340 of 7318 non-T2DM patients were on statins (32.0%). The decision tree analysis, logistic regression analysis, and survival analysis indicated that statin use in T2DM patients age < 65 years was significantly associated (OR = 0.55, p < 0.01) with a decreased likelihood for all-cause mortality at one, six, and 12 months. Analysis after PSM found statin use in T2DM patients age < 65 was associated with non-significantly decreased odds for one, six, and 12 month mortality (OR = 0.70, p = 0.09). The logistic regression analysis and PSM analysis showed no significant difference in mortality likelihood between T2DM patients age ≥ 65 with statin use and those without statin use. No significant difference was seen in rehospitalization between T2DM cases and controls in either age groups. CONCLUSIONS: Decision tree analysis, logistic regression analysis, and Cox regression analysis showed that the T2DM and statin interaction was significantly associated with decreased mortality at one, six, and 12 months for T2DM patients age, but not in T2DM patients age ≥ 65. A prospective case control study with a larger sample size to account for PSM may validate these findings to be significant. This dissertation emphasized the continued study of statin therapy for the attenuation of CAP severity and improved outcomes.