Aldosterone/direct renin concentration ratio as a screening test for primary aldosteronism: A meta-analysis

Abstract

Objective:The accuracy of aldosterone/direct renin concentration ratio (ADRR) as a screening test in patients with primary aldosteronism (PA) varies widely across the studies. Therefore, we conducted a meta-analysis to assess the accuracy of ADRR.Methods:A literature search was performed in PubMed, Embase, and the Cochrane library published between April 1971–February 2016. Studies focusing on the accuracy of ADRR for PA screening were included. Two authors independently extracted information regarding patient characteristics, antihypertensives status, true positives, true negatives, false positives, and false negatives. The random-effects model was used for statistical analysis. Heterogeneity was explored by subgroup analysis and meta-regression.Results:Nine studies involving 974 patients were included. The overall sensitivity, specificity, area under the curve, and diagnostic odds ratio of ADRR were 0.89 (95% confidence interval (CI) 0.84–0.93), 0.96 (95% CI 0.95–0.98), 0.985 and 324 respectively, with substantial heterogeneity. Meta-regression showed that antihypertensive status affects the ADRR and may account for the heterogeneity (p=0.03). Subgroup analysis of patients who discontinued the antihypertensives revealed a sensitivity of 0.99 (95% CI, 0.95–1.00) and a specificity of 0.98 (95% CI, 0.96–0.99).Conclusions:This study demonstrates the efficacy of ADRR as a screening test for PA. However, as antihypertensive drugs can interfere with the interpretation of ADRR, it is recommended to interrupt therapy or at least replace with analogues that do not significantly affect the ADRR value.

Age-Related Cutoffs of Plasma Aldosterone/Renin Concentration for Primary Aldosteronism Screening

“…Previous studies have revealed that PRA and PRC were well correlated and that the automated renin chemiluminescent assay was a reliable alternative to the radioimmunometric method [ 5 – 8 ]. Our previous meta-analysis that analyzed the accuracy of PAC/PRC as a screening test in patients with PA revealed that the overall sensitivity, specificity, and area under the curve of PAC/PRC were 0.89, 0.96, and 0.985, respectively, demonstrating the efficacy of PAC/PRC as a screening test for PA [ 9 ]. Currently, there are no unanimous cutoff values for the ARR for PA screening, and the PRC-based ARR cutoffs mentioned in previous studies varied between 1.0–5.7 (ng/dl)/( μ IU/ml) [ 3 – 8 , 10 – 14 ].…”

Clinical biomarker innovation: when is it worthwhile?

“…Because prevalence estimates differ across healthcare settings, we varied this parameter in a sensitivity analysis to investigate the relationship between PA prevalence and the headroom for a diagnostic test (see Analyses) [24]. For the ARR test, we assume a sensitivity of 0.89 and a specificity of 0.96 based on a meta-analysis conducted by Li et al [15]. For confirmed PA patients we assume a probability of 0.47 of having an aldosterone-producing adenoma that qualifies for surgery from Shah and Deshpande [25].…”

“…The results of the PSA are reported using the percentile method, yielding ranges similar to 95% confidence intervals (CIs). Furthermore, a univariate sensitivity analysis was conducted to investigate the individual impact of the following parameters: the sensitivity and specificity of the ARR (varied from 0.22 to 1 and 0.56 to 1, respectively, based on the ranges reported in a meta-analysis by Li et al [15]), the prevalence of PA (varied from 5% to 25%), and the cost-effectiveness threshold (varied from €20,000 to €80,000). The ARR sensitivity and specificity values were selected because of the heterogeneous figures reported in the literature and the PA prevalence rate and cost-effectiveness threshold based on potential differences across health care settings.…”

“…To exemplify, a novel PA test for hypertensive patients with a sensitivity of 0.90 and a specificity of 1 may cost up to €51 more than the current test and still be considered cost-effective. Note that for the current test we assumed a sensitivity of 0.89 and a specificity of 0.96 [15]. Figure 4 shows that a novel test for diagnosing PA may be worth the extra expenditure when its sensitivity is at least 0.90 and its specificity is at least 0.70.…”

“…Second, for reliable use of the ARR as a screening test -the Japan Endocrine Society even recommends screening all hypertensive patients for PA -the test's sensitivity should be very high to avoid false negatives [14]. Doubts regarding the ARR's suitability are evident from the many studies that investigated its diagnostic properties, indicating a repeated desire to evaluate the use of ARR and its cut-off value [15]. Third, optimization of the PA diagnosis is an active field of research, implying concerns regarding current diagnostic methods.…”