Description
TitleImpact of socio-demographic factors on comorbidity burden in diabetes mellitus patients
Date Created2015
Other Date2015-10 (degree)
Extent1 online resource (xiv, 303 p. : ill.)
DescriptionBACKGROUND — Racial/ethnic and Gender disparities in healthcare quality have been documented, but knowledge of discrepancy in diabetic comorbidities, prevalence, awareness and impact of sociodemographic factors among diabetic cohort remains incomplete, which may reflect direct effects of comorbidities and, medical care quality. Diabetes mellitus (DM) epidemic is relentless and a common metabolic disorder with major public health impact due to its detrimental consequences causing severe end-organ damage as retinopathy, nephropathy, cardiovascular, neurological and many more disorders. This study assessed the nationwide comorbidity patterns in multiracial/multiethnic hospitalized cohort of diabetic patients. OBJECTIVE — The purpose of this study was to assess the variability in expression, measure the discrepancies in prevalence rates, and probability of comorbid disorders associated with T1 and T2 diabetes independently, due to theoretically-empirically (sociodemographic, SDRFs) derived risk factors with in a diabetic cohort of hospitalized patients. This can be a vital attribute for health care providers that may serve as basis to reduce this gap of complications in the future, during a period when the prevalence of diabetes is predicted to increase dramatically. RESEARCH DESIGN AND METHODS — Data were obtained from all US states contributed to the Nationwide Inpatient Sample (NIS). All patients admitted to hospitals between 2004 and 2008 with a discharge diagnosis of T1 and T2 diabetes mellitus (DM, identified by the International Classification of Diseases, Ninth Revision procedure codes). Hypotheses testing consisted of both association and logistic regression analyses. In addition to a variety of summary statistics, bivariate analyses, multiple regression, two-sample ANOVA tests, t-tests, and chi-square were used to explore relationships within the data. Comparisons for patient’s demographics, and top comorbid diseases / conditions based on ICD-9-CM codes were done for the obese vs. the non-obese population. Multivariable Logistic regression modeling with variance; maximum R2, and ROC/AUC © in SAS 9.2 was used. RESULTS — Totals of 25,511 symptomatic T1D and 183,127 symptomatic T2D records were extracted between years 2004 and 2008 for use in this study. Overall T1D hospitalizations declined from 8,090 (18%) in 2004 to 3907 (8.2%) in 2008, whereas percentage of T2D hospitalizations due to comorbidities inclined from 31,981 (71.2%) in 2004 to 39,845 (83.3%) in 2008. Chi-square results of T1D depicted 23% nonobese males and 16% nonobese females significantly (<0.0001) developed comorbidities after adjusting the age group between 26-50 years (40%), while 27.7% of T2D nonobese males and 22.7% females affected with comorbid disorders (p<0.0001) after controlling the age between 51-75 years (51.6%). Two sample T test revealed the mean age in years of T1D men were lower than women (p<0.0001) and mean LOS in days for men were more than women (p=0.01). Mean age for T1D obese patients was higher than nonobese (p=0.02), but obesity had no significant (p=0.18) impact on LOS. Likewise, T2D men also had early onset of comorbid disorders than women (61, 65.5; p<0.0001) and LOS was also more for men (5.3; p<0.0001). Impact of obesity was also significant with earlier (55.04 years; p<0.0001) onset of comorbidities than non obese, but slightly lesser LOS than nonobese T2D patients. Two-way ANOVA test results disclosed the gender and racial disparity. Earliest onset of comorbidities was seen in black and NA T1D with LS-mean age of 45.8 years compared to late onset in asian/PIs (52 years). T2D race/ethnicity, gender, obesity and their interactions had significant impact on both LS-mean ages as wells LOS. Tukey’s studentized range /HSD multiple comparison tests revealed this age disparity due to significant differences among three interaction pairs for each race/ethnicity*gender (asian/PIs-white females, asian/PIs-white males-black-hispanic-NA females and black-hispanic-NA males) and race/ethnicity*obesity interaction (asian/PIs/white-non obese, Hispanic/black/NA-nonobese-white/Asian/PIs-obese and Hispanic/NA/black-obese patients). Black obese T2D patients had earliest onset of comorbid disorders. In multivariable logistic regression (MLR) analyses, significant disparities among comorbidities associated with diabetes mellitus (CADM) were found due to both theoretically (age, race/ethnicity, gender and obesity) and empirically-derived risk factors (length of stay, health insurance, median household income and patient’s location). Racial and gender comparisons revealed that T1D obese male patients had higher odds of dyslipidemia (OR; 1.2, 95% CI; 0.84-1.65; p<0.0001) and musculoskeletal disorders (OR; 1.3, 95% CI; 0.91-1.74; p< 0.0001) in Hispanics, systemic infection (OR; 1.3, 95% CI; 0.91-1.74; p< 0.0001) in asians/PIs, noncompliance (OR; 1.2, 95% CI; 0.86-1.61; p; 0.003) in native Americans , and white majority had recreational drug addiction (OR; 1.5, 95% CI; 1.4-1.8; p<0.0001). T2D obese males had higher odds of having systemic infection (OR; 1.7, 95% CI; 1.6-1.8; p; 0.0002) in asian/PIs and Hispanics, while noncompliance (OR; 2.4, 95% CI; 2.2-2.7; p<0.0001) and recreational drug addiction (OR; 1.4, 95% CI; 1.3-1.5; p<0.0001) was primarily seen in black minorities. Conclusively, all sociodemographic risk factors were independently linked to the elevated burden of comorbidities among inpatient cohort of diabetic patients. CONCLUSIONS — This study confirms significant correlation between CADM and sociodemographic (theoretically-empirically-derived) risk factors: age, race/ethnicity, gender, obesity, length of hospital stay, type of health insurance, patient’s median house hold income and location. In logistic regression analysis, the variables; age group, race/ethnicity, gender and health insurance explained a significant portion of the variance in development of comorbidities in this diabetic cohort sample. The persistence of ethnic disparities, in particular, after adjustment suggests a possible genetic origin, the contribution of unmeasured environmental factors, or a combination of these factors. This study contributed to the body of knowledge regarding in-hospital comorbidity burden in diabetic patients due to disparities among socio demographic characteristics and also provided new evidence that rates of selected comorbidities were similar or higher/lower compared to white majority, however, more empirical evidence is needed to further validate these specific individual/hospital-level characteristics that may be targets for facilitating improvements in the diabetic population as there are probably other factors which are not included in the NIS database.
NotePh.D.
NoteIncludes bibliographical references
Noteby Sheelu Bhatnagar
Genretheses, ETD doctoral
Languageeng
CollectionSchool of Health Related Professions ETD Collection
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.
Statistical Profile