Browsing by Subject "Medication adherence"
Now showing 1 - 4 of 4
Results Per Page
Sort Options
Item Depression in patients with diabetes : risk factors, medication-taking behaviors, and association with glycemic control(2010-12) Suehs, Brandon Thomas; Lawson, Kenneth Allen, 1952-; Smith, Tawny Bettinger; Barner, Jamie C.; Crismon, Miles L.; Garcia, Alexandra A.This study evaluated the epidemiological relationship between diabetes and incident depression, as well as antidepressant medication utilization among indigent care patients diagnosed with diabetes. Medical data for 2,886 subjects receiving care in a public indigent care provider network were utilized for this study. Diagnoses of diabetes, depression, and other comorbid medical conditions were identified from the electronic medical record. Prescription claims data from the clinic pharmacy network were used to evaluate medication-taking behaviors. Clinical laboratory data were extracted, as available, from the electronic clinic records. After controlling for the influence of age, gender, race/ethnicity, marital status classification, and Charlson score, a diagnosis of diabetes was associated with a 42 percent reduction in odds of new-onset depression (p = 0.021). In the a priori analysis of factors associated with new-onset depression among diabetic patients, none of the risk factors evaluated were associated with incident depression at a statistically significant level. Post-hoc exploratory analyses revealed that female gender and White non-Hispanic race/ethnicity were associated with increased odds of a prevalent diagnosis of depression among diabetic patients. Patients with diabetes were more likely to be prescribed selective serotonin reuptake inhibitors (SSRIs) as their initial antidepressant medication compared to non-SSRIs. Diagnosis of diabetes was not associated with antidepressant switch, discontinuation, or 6-month antidepressant adherence; however, diagnosis of diabetes was associated with a higher level of 12-month antidepressant adherence (p = 0.024). Diagnosis of diabetes was also associated with a higher level of 3-month antidepressant persistence (p = 0.004), but not 12-month persistence. There were no statistically significant relationships observed between initial class of antidepressant medication prescribed and any of the medication-taking behaviors evaluated. For subjects with available data (n = 106), glycemic control was evaluated in terms of hemoglobin A1c. Increased antidepressant medication adherence was associated with higher hemoglobin A1c values during follow-up. Results suggest that prevalent diabetes is associated with a reduced risk of diagnosis of new-onset depression in indigent care patients. Further research is necessary to evaluate the effect that chronic comorbid medical conditions such as diabetes may have on antidepressant medication-taking behaviors, and the relationship between antidepressant exposure and glycemic control.Item Long-term adherence and outcomes of oral Tyrosine Kinase Inhibitors for the treatment of CML in the US VHA medical system(2014-08) Kreys, Eugene Daniel; Koeller, Jim; Frei, Christopher R; Wilson, James P; Shepherd, Marvin D; Bollinger, Mary JChronic Myeloid Leukemia (CML) represents about 15-20% of all adult leukemias. The introduction of Tyrosine Kinase Inhibitors (TKIs) was a breakthrough in the treatment of CML that drastically improved outcomes. Poor adherence is recognized to be a major source of treatment failure and is especially concerning in situations where medications are self-administered, as is the case with TKI therapy. Several published studies on patient adherence with oral chemotherapy found rates for long-term treatment to be around 40-50%. The primary purpose of this study was to determine long-term adherence to TKI therapy, and to establish the effect of adherence on the clinical response. A secondary purpose was to compare adherence and treatment outcomes among TKIs. This was a retrospective cohort study of CML patients receiving TKI therapy at any Veteran Health Administration (VHA) facility. Patients 18-89 years of age, with CML diagnosis that filled at least one prescription for imatinib, nilotinib, or dasatinib from 10/1/2001 through 9/30/2010 were included in this study. Adherence was ascertained for 2,873 patients by calculating the Medication Possession Ratio (MPR) using administrative refill data. A manual chart review of 683 patients determined the clinical effectiveness of TKI therapy by identifying cases of major molecular response (MMR), as well as complete cytogenetic response (CCyR). Thirty-three percent of dasatinib-treated patients were adherent during first-year of treatment relative to 28% of nilotinib-treated patients, resulting in an adjusted OR 1.24 (95% CI: 0.78-1.95, p= 0.361). Fifty-one percent of the patients receiving dasatinib as second-line treatment achieved documented MMR by 18 months relative to 56% of nilotinib-treated patients, resulting in an adjusted OR of 0.66 (95% CI: 0.35 -1.23, p= 0.189). Documented MMR by 18 months was achieved by 53% of the patients adherent to TKI therapy relative to 45% of nonadherent patients. When adjusted for covariates, the difference was significant with an OR of 2.68 (95% CI: 1.58 - 4.57, p< 0.001). In conclusion, no significant difference in adherence rates or clinical effectiveness was observed between dasatinib or nilotinib when administered as second-line treatment. Adherence to TKI therapy was found to be significantly associated with improved clinical effectiveness.Item Medication adherence, persistence, switching and dose escalation with the use of tumor necrosis factor (TNF) inhibitors among Texas Medicaid patients diagnosed with rheumatoid arthritis(2013-08) Oladapo, Abiola Oluwagbenga; Barner, Jamie C.The main purpose of this study was to evaluate medication use patterns (i.e., dose escalation, medication adherence, persistence, and switching) of rheumatoid arthritis (RA) patients on etanercept (ETN), infliximab (IFX) or adalimumab (ADA) and the associated healthcare utilization costs using Texas Medicaid data. Study participants were Medicaid beneficiaries (18-63 years) with an RA diagnosis (ICD-9-CM code 714.0x) who had no claim for a biologic agent in the 6-month pre-index period (July 1, 2003 - Dec 31, 2010). The index date was the first date when the patient had the first fill for any of the study TNF inhibitors (ETN, ADA or IFX) within the study identification period (Jan 1, 2004 – Aug 31, 2010). Data were extracted from July 1, 2003 to August 31, 2011. Prescription and medical claims were analyzed over an 18-month study period (i.e., 6-month pre-index and 12-month post-index periods). The primary study outcomes were adherence, persistence, dose escalation, switching and cost (i.e., total healthcare, RA-related and TNF inhibitor therapy cost). The study covariates were demographic factors (age, gender, race/ethnicity), pre-index use of other RA-related medications (pain, glucocorticoids and disease modifying antirheumatic drugs), total number of non-study RA-related medications used at index, pre-index RA and non-RA related visits, pre-index healthcare utilization cost and Charlson Comorbidity Index score. Conditional regression analyses, which accounts for matched samples, were used to address the study objectives. After propensity score matching, 822 patients (n=274/group) comprised the final sample. The mean age (±SD) was 48.9(±9.8) years, and the majority of the subjects were between 45 and 63 years (69.2%), Hispanic (53.7%) and female (88.0%). Compared to patients on ETN, the odds of having a dose escalation were ≈ 5 [Odds Ratio= 4.605 [95% CI= 1.605-12.677], p=0.0031] and ≈ 8 [Odds Ratio=7.520, [95% CI= 2.461-22.983], p=0.0004] times higher for IFX and ADA patients, respectively, while controlling for other variables in the model. Compared to ETN, patients on IFX (p=0.0171) were more adherent while adherence was comparable with patients on ADA (p=0.1144). Compared to patients on ETN, the odds of being adherent (MPR ≥ 80%) to IFX was ≈ 2 times higher [Odds Ratio= 2.437, [95% CI=1.592-3.731], p < 0.0001] while controlling for other variables in the model. Persistence to index TNF inhibitor therapy and likelihood to switch or discontinue index TNF inhibitor therapy were comparable among the 3 study groups. In addition, the duration of medication use (i.e., persistence) prior to switching or discontinuation of index therapy was comparable among the 3 study groups. Furthermore, for each of the cost variables (total healthcare, RA-related and TNF inhibitor therapy cost), costs incurred by patients on ETN were significantly lower (p < 0.01) than those incurred by ADA patients but significantly higher (p < 0.01) than those incurred by IFX patients. Finally, a positive and significant relationship (p < 0.0001) was found between RA-related healthcare cost, adherence and persistence to TNF inhibitor therapies. In conclusion, ETN was associated with lower rates of dose escalation compared to ADA or IFX. However, adherence was better and associated healthcare costs were lower with IFX. Clinicians should endeavor to work with each individual patient to identify patient-specific factors responsible for poor medication use behaviors with TNF-inhibitor therapies. Reducing the impact of these factors and improving adherence should be included as a major part of the treatment plan for each RA patient. RA patients need to be adequately educated on the importance of adhering and persisting to their TNF-inhibitor therapy as poor medication adherence/persistence negatively impacts the RA disease process.Item Outcome prediction and structure discovery in healthcare data(2016-08) Arzeno-González, Natalia María; Vikalo, Haris; Ghosh, Joydeep; Lawson, Karla A; Sanghavi, Sujay; Vishwanath, SriramGrowing use of electronic medical records, advances in data mining and machine learning, and the continually increasing cost of healthcare in the United States drive the necessity of algorithmic solutions with the potential to improve patient care and reduce healthcare costs. Such algorithms can enable the identification of the most relevant parameters for predicting adverse events, reveal underlying physiological mechanisms of diseases, and determine likelihood of complications that may lead to rehospitalization of discharged patients. Key limitations in computational tools currently used in healthcare or with the potential to greatly benefit the healthcare system can be overcome by methods that allow for soft constraints or promote smoothness. In this dissertation we develop three main algorithms incorporating softness or smoothness in the constraints or solution and demonstrate applications in diverse aspects of healthcare with the potential to greatly reduce healthcare costs. We first develop an outcome prediction algorithm that preserves the clinical knowledge from the development of additive risk scores with hard thresholds (of the form add p points if variable x is above/below threshold t). This novel method is not only easily optimizable for different patient sub-populations, but reveals clinically interpretable information such as the maximum contribution of a physiologic variable to the risk score and the range of values for which risk increases. We then turn to overcoming limitations in two clustering settings. In a semi-supervised setting, where pairwise constraints (relationships between pairs of points) are available, we develop an algorithm capable of performing accurate clustering under noisy constraints. This is achieved via soft constraints that impose a penalty on the objective when violated. Finally, we examine the scenario where clustering data are available at multiple points in time under the assumption of temporal smoothness, i.e., data points are more likely to remain in the same cluster than to change cluster membership between consecutive time steps. In this setting, we develop an evolutionary clustering algorithm that automatically infers the number of clusters at each time and matches the clusters across time steps while finding a global clustering solution. The proposed schemes outperform existing methods in benchmark and non-healthcare datasets as well as in the tasks of mortality prediction from clinical data and breast cancer metastasis prediction from gene expression data. As an additional healthcare application, we use our proposed evolutionary clustering algorithm to study the evolution of health plan clusters inferred from medication adherence data and provide a detailed analysis of the clusters.