Hypertension affects one third of the U.S. adult population. Although cost-effectiveness analyses of antihypertensive medicines have been published, a comprehensive systematic review across medicine classes is not available. PubMed, Embase, Cochrane Library, and Health Technology Assessment were searched to identify original cost-effectiveness analyses published from 1990 through August 2016. Results were summarized by medicine class: angiotensin-converting enzyme inhibitors (ACEIs), angiotensin receptor blockers (ARBs), calcium channel blockers (CCBs), thiazide-type diuretics, β-blockers, and others. Incremental cost-effectiveness ratios (ICERs) were adjusted to 2015 U.S. dollars. Among 76 studies reviewed, 14 compared medicines with no treatment, 16 compared medicines with conventional therapy, 29 compared between medicine classes, 13 compared within medicine class, and 11 compared combination therapies. All antihypertensives were cost effective compared with no treatment (ICER/quality-adjusted life year [QALY]=dominant-$19,945). ARBs were more cost effective than CCBs (ICER/QALY=dominant-$13,016) in nine comparisons, whereas CCBs were more cost effective than ARBs (ICER/QALY=dominant) in two comparisons. ARBs were more cost effective than ACEIs (ICER/QALY=dominant-$34,244) and β-blockers (ICER/QALY=$1,498-$18,137) in all eight comparisons. All antihypertensives were cost effective compared with no treatment. ARBs appeared to be more cost effective than CCBs, ACEIs, and β-blockers. However, these latter findings should be interpreted with caution because these findings are not robust due to the substantial variability across the studies, including study settings and analytic models, changes in the cost of generic medicines, and publication bias.

The purpose of this study is to investigate comorbidity status and its impact on total medical expenditures in non-institutionalized hypertensive adults in the U.S. Data from the 2011-2014 Medical Expenditure Panel Survey were used. Patients were included if they had a diagnosis code for hypertension, were aged ≥18 years, and were not pregnant during the study period (N=26,049). The Elixhauser Comorbidity Index was modified to add hypertension-related comorbidities. The outcome variable was annual total medical expenditures, and a generalized linear model regression (gamma distribution with a log link function) was used. All costs were adjusted to 2014 U.S. dollars. Based on the modified Elixhauser Comorbidity Index, 14.0% of patients did not have any comorbidities, 23.0% had one, 24.4% had two, and 38.7% had three or more. The five most frequent comorbidities were hyperlipidemia, diabetes, rheumatoid arthritis, depression, and chronic pulmonary disease. Estimated mean annual total medical expenditures were $3,914 (95% CI=$3,456, $4,372) for those without any comorbidity; $5,798 (95% CI=$5,384, $6,213) for those with one comorbidity; $8,333 (95% CI=$7,821, $8,844) for those with two comorbidities; and $13,920 (95% CI=$13,166, $14,674) for those with three or more comorbidities. Of the 15 most frequent comorbidities, the condition with the largest impact on expenditures for an individual person was congestive heart failure ($7,380). Hypertensive adults with stroke, coronary heart disease, diabetes, renal diseases, and hyperlipidemia had expenditures that were $6,069, $6,046, $5,039, $4,974, and $4,851 higher, respectively, than those without these conditions. Comorbidities are highly prevalent among hypertensive adults, and this study shows that each comorbidity significantly increases annual total medical expenditures.

There is a lack of research focused on understanding the different characteristics and healthcare utilization of metastatic breast cancer patients by palliative care use. This study aims to investigate trend of in-patient palliative care and its association with healthcare utilization among hospitalized metastatic breast cancer patients in the US. National Inpatient Sample (NIS) was used to identify nationwide metastatic breast cancer patients (n = 5209, weighted n = 25,961) from 2010 to 2014. We examined the characteristics of the study sample by palliative care and its association with healthcare utilization, measured by discounted hospital charges and length of stay. Multivariable survey regression models were used to identify predictors. Among 26,961 breast cancer patients, 19.0% had palliative care. Percentage of receiving palliative care during the period were gradually increased. Social factors including race, insurance types were also associated with a receipt of palliative care. Survey linear regression results showed that patients with palliative care were associated with 31% lower hospital charges, however, length of stays were not significantly associated. This study found evidence of who was associated with the receipt of palliative care and its relationship with healthcare utilization. This study also emphasizes the importance of receiving palliative care in patients with breast cancer, paving the way for future research into ways to improve palliative care in cancer patients. This study also found social differences and gave evidence of programs that could be used to help vulnerable groups in future health policy decisions.

Cardiovascular disease (CVD) in cancer patients can affect the risk of unplanned readmissions, which have been reported to be costly and associated with worse mortality and prognosis. We aimed to demonstrate the feasibility of using machine learning techniques in predicting the risk of unplanned 180-day readmission attributable to CVD among hospitalized cancer patients using the 2017–2018 Nationwide Readmissions Database. We included hospitalized cancer patients, and the outcome was unplanned hospital readmission due to any CVD within 180 days after discharge. CVD included atrial fibrillation, coronary artery disease, heart failure, stroke, peripheral artery disease, cardiomegaly, and cardiomyopathy. Decision tree (DT), random forest, extreme gradient boost (XGBoost), and AdaBoost were implemented. Accuracy, precision, recall, F2 score, and receiver operating characteristic curve (AUC) were used to assess the model’s performance. Among 358,629 hospitalized patients with cancer, 5.86% (n = 21,021) experienced unplanned readmission due to any CVD. The three ensemble algorithms outperformed the DT, with the XGBoost displaying the best performance. We found length of stay, age, and cancer surgery were important predictors of CVD-related unplanned hospitalization in cancer patients. Machine learning models can predict the risk of unplanned readmission due to CVD among hospitalized cancer patients.