* Candidate for the Lee B. Lusted Student Prize Competition
Purpose: Cholesterol lowering with statins and fibrates is a cornerstone of preventive cardiology. There is, however, little data on the magnitude of reduction in different endpoints in various high risk populations. In this analysis, we sought to determine how cholesterol lowering lowers the risk of cardiovascular death (CV death), myocardial infarction (MI) and stroke (CVA) in different populations.
Method: We performed a literature search to identify all clinical trials from 1994 onward for cholesterol lowering with statins or fibrates. Thirty two studies were identified and then segregated into 5 separate categories based on the underlying cardiovascular risk: Primary prevention (N=6), secondary prevention (N=8), comparison of drug dose in secondary prevention (N=8), significant comorbid conditions (N=6), and fibrates (N=4). Separate meta-analyses were performed for the three different endpoints for each of these 5 categories.
Result: Primary and secondary prevention populations had significant reduction in all categories: CV death (25 and 22% reduction respectively), MI (41% and 32% reduction), and CVA (23% and 21% reduction). High dose versus low dose statins did not significantly decrease CV death, but had continued further significant reduction in MI (22%) and CVA (13%). Treatment with statins in comorbid diseases, such as heart failure and renal failure did not affect CV death or CVA, but did lower risk of MI by 23%. Fibrate therapy, based on these 4 studies, did not significantly affect CV death or CVA, but had a borderline significant decrease in MI of 18%. Rates for CV death ranged from 0.3% per year for primary prevention to 5.15% with cormorbidities. MI rates were similar at approximately 1.1% per year in all groups except primary prevention (0.56%). CVA rates were lowest in primary prevention (0.46% and highest in patients with comorbidities (1.2%).
Conclusion: This data suggests that while there is a virtually uniform relative risk reduction of MI with cholesterol lowering with statins or fibrates, this effect is not necessarily carried over to reduction in CVA or CV death. The underlying comorbidities, disease processes, and the extent of relative and absolute risk reduction in each of these three endpoints need to be assessed separately in defining the expected benefit of cholesterol reduction in diverse populations.
Purpose: To formulate best estimates of sensitivity and specificity of commonly used biomarkers for identifying Alzheimer disease (AD) amongst dementia patients.
Method: Mild AD is often difficult to differentiate from mild cognitive impairment or non-AD dementias. However, making the distinction of AD is an important step to guide therapy and advise patients and caregivers. Suggested revisions to the NINDS-ADRDA criteria for diagnosing AD call for addition of one or more biomarkers. The tangles and plaques observed in AD patients are formed by excess tau filaments and amyloid protein deposits. Increased concentrations of neurofillaments and decreased concentrations of amlyoid in the cerebrospinal fluid (CSF) may be characteristic of AD. Total tau, phosphorylated tau (Ptau), and 42 amino acid form of beta-amyloid (Aβ1-42) concentration in the CSF have been studied as biomarkers for AD. A search was conducted to locate all studies of biomarkers for AD published in English from January, 1990 to March 2010. Synthesis was performed using a bivariate mixed-effects binary regression model. We calculated sensitivity (SN), specificity (SP) and the receiver operating curves (ROC), with confidence and prediction contours.
Result: Of 1840 unique studies identified, 20 presented primary data sufficient for analysis. By testing levels of tau in CSF, AD could be discriminated from non-AD dementias with a SN of 75% (95%CI 70% to 80%) and SP of 77% (95%CI 70% to 82%). The level of Ptau in CSF had a SN of 75% (95%CI 66% to 82%) and SP of 82% (95%CI 72% to 89%). The level of Aβ1-42 had a SN of 72% (95%CI 66% to 77%) and SP of 67% (95%CI 61% to 72%), while tests combining the levels of tau and Aβ1-42 resulted in SN of 80% (95%CI 72% to 85%) and SP of 76% (95%CI 57% to 88%). The area under the ROC was highest for the test of Ptau, 85% (95%CI 82% to 88%). The proportion of heterogeneity likely due to the threshold effect ranged from 17% for tau to 100% for the combined tau and Aβ1-42.
Conclusion: A combination test for Aβ1-42 and tau resulted in the highest SN, but the Ptau test had the highest SP and AUROC. Treatment profiles will determine whether more accurate SN or SP has a higher value in AD diagnosis.
Purpose: With the development of highly active anti-retroviral therapy (HAART) to treat HIV infection, HAART managed persons can now be expected to live longer than those in the pre-HAART era. As a result, many persons living with HIV infection will now experience end-stage organ failure (ESOF) well before they have life-threatening conditions related to this disease. Given the improved prognosis for people living with HIV infection and the burden of illness they may experience from ESOF, the benefit of solid organ transplantation to treat ESOF in this population needs to be determined. The purpose of this systematic review and meta-analysis was to determine the effectiveness of solid organ transplantation in persons with end stage kidney and liver failure and human immunodeficiency virus (HIV+).
Method: Multiple bibliography databases (e.g. OVID MEDLINE, EMBASE) were searched (“HIV”, “Liver, Kidney Transplantation) for potentially relevant citations (n=1204) from 1996 to 2009. Identified studies were full-text reviewed according to the pre-defined inclusion criteria i) RCT, observational studies, ii) HIV+ population receiving HAART with ESOF, iii) studies reporting Kaplan-Meier survival curve analysis, iv) minimum of 1-year follow, and v) English language. Study characteristics and survival curve data were extracted by one reviewer and verified by another. Quality of evidence was assessed using GRADE. Hazard ratio (HR) estimates were constructed from summary survival data, if necessary; they were combined into pooled HRs using random-effects models for the following outcomes: i) death after transplantation, ii) graft survival iii) HIV disease progression, iv) acute graft rejection, and v) recurrence of hepatitis C (HCV) infection.
Result: Fifteen studies were identified after full-text review. The quality of the evidence is very low. The risk of death after kidney transplantation does not differ between HIV-positive and HIV-negative study participants (pooled HR: 0.90 [95% CI: 0.36, 2.23]). The risk of death after liver transplantation is statistically significantly 64% higher for HIV-positive than HIV-negative participants (pooled HR 1.64 [1.32, 2.02]). The risk of death after liver transplantation is statistically significantly 2.8 fold greater in HIV/HCV co-infected participants than in HCV-positive mono-infected participants (pooled HR 2.81 [1.47, 5.37]).
Conclusion: Outcomes among HIV infected kidney transplant patients are not worse than uninfected patients; outcomes among liver transplantation patients appear to be worse. These conclusions are uncertain because of very low quality evidence.
Purpose: There have been very few randomized controlled trials (RCTs) comparing Laparoscopic Nissen Fundoplication (LNF) to medical management using proton pump inhibitors (PPIs) for treating patients with chronic gastro-esophageal reflux disease (GERD). These RCTs have reported mixed results with respect to symptom control and impact on patient quality of life (QOL) and economic evaluations of LNF have been inconclusive. The objective of this study is to determine the incremental cost-utility (CU) of LNF versus PPI for the treatment of patients with chronic GERD over a three-year period from the societal perspective.
Methods: The economic evaluation was conducted alongside a RCT in Hamilton, Canada. This RCT enrolled 104 patients from October 2000 to September 2004 and followed them for three years. The primary study outcome was GERD symptoms and secondary outcome measures included QOL and cost-utility (CU). Health resource utilization (total and GERD-related only) and QOL data (Health Utilities Index (HUI3), Short Form 6D (SF6D) and EQ5D ) were collected at regular follow-up intervals to determine the incremental cost per quality-adjusted-life-year (QALY) gained. Stochastic uncertainty was assessed using bootstrapping and methodological assumptions were assessed using traditional deterministic sensitivity analysis.
Results: There were no statistically significant differences in GERD symptom scores over 3 years, but LNF did result in fewer heartburn (HB) days and improved QOL. Costs were higher for LNF patients by a mean of $3,205 per patient over 3 years but QOL was also higher as measured by any QOL instrument. Based on total costs, the incremental CU of LNF was $29,404 per QALY gained when using the HUI3 and stochastic analysis of uncertainty showed LNF to be cost-effective at willingness-to-pay (WTP) ceiling ratio thresholds above $30,000 per QALY. CU results were sensitive to the utility instrument used ($29,404 per QALY for HUI3, $31,117 per QALY for SF6D and $76,310 per QALY for EQ5D).
Conclusions: Results varied depending on resource use/costs included in the analysis and the QOL instrument used, however, LNF was generally found to be a cost-effective treatment for symptomatic GERD patients requiring long-term management.
Purpose: Patients with chronic kidney disease may not consistently receive optimal care and chronic care models that rely on physicians or nurses that closely track disease progression may be beneficial in reducing the rate of disease progression. This study estimated the cost effectiveness of a chronic disease management model for chronic kidney disease.
Methods: In a multi-centre randomized control trial, the Canadian Prevention of Renal and Cardiovascular Endpoints Trial (CanPREVENT), patients with chronic kidney disease received either usual care or a nurse/nephrologist supported chronic care model that targeted factors associated with development of kidney and cardiovascular disease events. Cost and outcomes from CanPREVENT were compared to determine the incremental cost-effectiveness of the chronic disease management model with probabilistic analysis. Base case analysis included disease-related costs, and sensitivity analysis included all costs recorded for each subject.
Results: 238 patients received the nurse/nephrologist intervention and 236 received the control usual care. Over 2 years, there was a trend towards less disease-related costs in the intervention group (intervention: $4,631, control: $5,741: difference $1,109, P=0.146). All costs were significantly lower in the intervention group (intervention: $11,739, control $14,180, difference $2,441: P=0.023). These differences were mostly related to lower hospitalization costs. In addition, average HUI3 utility score increased by 0.024 in the intervention group, while declining by 0.021 in controls group patients over two years P=0.013. In the base case analysis and with all costs, the intervention dominates the control group. The results are robust to including disease costs only versus all costs, payer versus societal perspective, and to changes in the discount rate.
Conclusions: CanPREVENT was conducted to investigate the feasibility, effectiveness and cost effectiveness of a nephrolgist/nurse based multifaceted intervention for patients with CKD. Based on these economic results, CanPREVENT represents good value for money because it reduces costs and improves outcomes.
Purpose: Recent diabetes trials evaluating very intensive glucose control have produced results that appear inconsistent with prior epidemiological studies. Forecasting models can help evaluate the relationships between epidemiological studies and trials. We utilize a forecasting model of diabetes complications to assess the predictability of recent trial results based on prior epidemiology.
Methods: Models derived from United Kingdom Prospective Diabetes Study (UKPDS) was[V1] used to make outcome predictions based on the trial designs and patient populations for the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial, Action in Diabetes and Vascular Disease (ADVANCE) trial, and Veterans Affairs Diabetes Trial (VADT). These predictions were then compared to the actual results. Predicted event rates (for both intensive and control arms) and relative risks were compared and judged on whether they fell within a calculated 95% confidence interval from the actual trials.
Result: For ADVANCE, the predicted event rates fell within the 95% confidence interval for 7 out of 18 compared outcomes. For VADT, this number was 5 out of 14. For ACCORD, model predicted event rates did not match any of the actual 12 event rates. The predicted relative risks fell within the 95% confidence interval for 8 out of 9 ADVANCE outcomes, 6 out of 7 VADT outcomes, and 2 out of 6 ACCORD outcomes.
Conclusion: The ACCORD trial represents a “black swan” trial, producing results that are unpredictable in relation to past diabetes epidemiology. These unpredictable results demand a renewed focus on the overall health effects of glucose lowering medication especially among the elderly and those with long-standing diabetes.