Key Points Question Are low levels of vitamin D independently associated with the risk of SARS-CoV-2 seropositivity? Findings In this cohort study of 18 148 individuals whose vitamin D levels were measured before the COVID-19 pandemic, low levels of vitamin D were associated with SARS-CoV-2 seropositivity in unadjusted univariable analysis. However, after adjusting for potentially confounding factors, including age, sex, race/ethnicity, education, body mass index, blood pressure, smoking status, and geographical location, vitamin D level was not associated with SARS-CoV-2 seropositivity. Meaning Although SARS-CoV-2–seropositive individuals did have lower vitamin D levels than seronegative individuals, low vitamin D levels were not independently associated with the risk of seropositivity.

The worldwide prevalence of type 2 diabetes (T2D) continues to increase, despite the established efficacy of T2D prevention interventions. Identifying individuals at high risk and making them the focus of preventive interventions may reduce the incidence of diabetes and global disease burden. We recently showed that T2D risk assessment in white middle-aged men and women can be improved with addition of an insulin resistance measure (assessed by HOMA of insulin resistance) to glycemia and other established risk factors (1). We have also developed an insulin resistance score (IRScore), comprising fasting insulin and C-peptide measured by mass spectrometry, to assess the probability of existing insulin resistance where insulin resistance was defined as being in the top tertile of steady-state plasma glucose level (≥198 mg/dL) (2). In the current study, we asked whether this IRScore improved T2D risk assessment beyond glycemia and established risk factors in a population of older Europeans. Older populations are of particular interest given the steady increase in the average age worldwide, especially in Europe, Japan, China, and the U.S. We conducted a case-cohort study based on the …

In clinical trials, vitamin D supplementation has been reported to reduce serum levels of total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and triglycerides (TG) but not high-density lipoprotein cholesterol (HDL-C). In this cohort study we evaluated the association between changes in vitamin D (25-hydroxyvitamin D) and changes in lipid levels in a real-world setting. Changes in lipid levels over a 1-year period were evaluated among individuals whose vitamin D levels increased (group 1) or decreased (group 2) by ≥ 10 ng/mL in year 2018 versus 2017 (cohort 1; n = 5580), in 2019 versus 2018 (cohort 2, n = 6057), or in 2020 versus 2019 (cohort 3, n = 7249). In each cohort, levels of TC, LDL-C, and TG decreased in group 1 and increased in group 2. Between-group differences in average changes in the 3 cohorts ranged from 10.71 to 12.02 mg/dL for TC, from 7.42 to 8.95 mg/dL for LDL-C, and from 21.59 to 28.09 mg/dL for TG. These differences were significant after adjusting for age, sex, race, education, body mass index, blood pressure, smoking status, geographical location, and baseline levels of vitamin D and lipids (P < 0.001). Changes in vitamin D levels were not significantly associated with changes in HDL-C levels.

Question What is the within-couple concordance of cardiovascular (CV) risk factors and behaviors? Findings In this cross-sectional study of 5364 couples consisting of employees and spouses (or domestic partners) undergoing an annual employer-sponsored health assessment, 79% of the couples were in the nonideal category of a CV health score. This within-couple concordance of nonideal CV health scores was associated mostly with unhealthy diet and inadequate physical activity. Meaning The high concordance of nonideal CV risk factors and behaviors within couples suggests that behavioral modification programs may benefit both the targeted and the nontargeted member of the couple. This cross-sectional study assesses the within-couple concordance of cardiovascular risk factors and behaviors within couples who participated in a nationwide US employer-sponsored health assessment program. Importance Married couples and domestic partners have been reported to share similar environmental exposures, adopt similar behavior patterns, and have similar transferable characteristics. However, the degree to which couples share similar levels of cardiovascular risk factors and behaviors is uncertain. Objective To assess within-couple concordance of the American Heart Association-defined Life's Simple 7 (LS7). Design, Setting, and Participants Cross-sectional study with a longitudinal substudy of employees and spouses (or domestic partners) who participated in an employer-sponsored health assessment program throughout the United States between October 2014 and December 2018. Data were analyzed from November 1, 2019, to August 4, 2020. Exposures Having a spouse or domestic partner. Main Outcomes and Measures The LS7 risk factors and behaviors (smoking status, body mass index, exercise, diet, total cholesterol, blood pressure, and fasting glucose) were assessed by questionnaires, examinations, and laboratory tests. LS7 categories were scored as 2 for ideal, 1 for intermediate, or 0 for poor and summed to generate a CV health score. Results The study included 10728 participants (5364 couples): 7% were African American, 11% Hispanic, 21% Asian, and 54% White (median [interquartile range] age, 50 [41-57] years for men and 47 [39-55] for women). For most couples, both members were in the ideal category or both were in a nonideal category. Concordance ranged from 53% (95% CI, 52%-54%) for cholesterol to 95% (95% CI, 94%-95%) for diet. For the CV health score, in 79% (95% CI, 78%-80%) of couples both members were in a nonideal category, which was associated mainly with unhealthy diet (94% [95% CI, 93%-94%] of couples) and inadequate exercise (53% [95% CI, 52%-55%] of couples). However, in most couples, both members were in the ideal category for smoking status (60% [95% CI, 59%-61%] of couples) and glucose (56% [95% CI, 55%-58%]). Except for total cholesterol, when 1 member of a couple was in the ideal category, the other member was likely also to be in the ideal category: the adjusted odds ratios for also being in the ideal category ranged from 1.3 (95% CI, 1.1-1.5; P <= .001) for blood pressure to 10.6 (95% CI, 7.4-15.3; P <= .001) for diet. Concordance differed by ethnicity, socioeconomic status, and geographic location. A 5-year longitudinal analysis of 2186 couples found modest changes in concordance of blood pressure (from 55% [95% CI, 53%-57%] to 59% [95% CI, 57%-61%]; P < .001 for trend) and fasting glucose (from 64% [95% CI, 62%-66%] to 59% [95% CI, 57%-61%]; P < .001 for trend) with no change in other factors. Conclusions and Relevance In this study, high concordance of nonideal behaviors was found within couples; behavioral modification programs may benefit both the targeted and the nontargeted member of a couple.

Introduction: Achieving optimal levels of low density lipoprotein cholesterol (LDL-C) reduces risk of cardiovascular events. We have found substantial gaps in LDL-C goal attainment in employees and spouses with an employer-sponsored medical plan. Understanding how patients’ engagement with healthcare providers affects LDL-C goal attainment may inform gap-closure programs. Methods: The study was drawn from 35,276 participants in an annual health assessment program offered by nation-wide clinical diagnostic provider to employees and spouses between September 2017 and June 2018. We excluded those with <1 consecutive year of employer- sponsored health plan and those younger than 40 or older than 75. The analysis was further limited to 7,363 participants who would benefit from LDL-C management according to the AHA/ACC definition of patient-management groups. The association between attaining LDL-C goal and education, socioeconomic status, and self-reported measures of engagement with healthcare providers was assessed in age, sex, and ethnicity adjusted logistic regression models. Results: Attaining LDL-C goal was more likely in those with a personal physician than in those without (OR =2.25, 95%CI, 1.70 to 2.99, P <0.0001). Similarly, greater confidence in selecting a physician, talking with a physician, or understanding medical benefits, as well as a more recent physical exam and a greater number of medications were associated with greater odds of attaining LDL-C goal ( Figure, all P values <0.0001). Socioeconomic status and education levels were not associated with attaining LDL-C goals ( P >0.05). Conclusion: Confidence in interaction with healthcare providers (understanding benefits, selecting a physician, and talking with a physician) was associated with LDL-C goal attainment in those with an employer-sponsored health plan. Improving patients’ confidence in their ability to interact with healthcare providers may improve LDL-C goal attainment.

We evaluated the cost-effectiveness of test-and-treat scenarios for vaginitis, scenarios based on clinical and microscopic examination (CME), nucleic acid amplification testing (NAAT), or nonamplified nucleic acid probe (probe) testing. The symptom resolution outcome and the payer cost of diagnosis and treatment were estimated in decision analytical models in a hypothetical patient population. Compared with probe testing, NAAT resulted in symptom resolution in more patients (615 versus 475 per 1000 tested) at a cost of $210 per incremental symptom resolution, a cost lower than the willingness to pay for symptom resolution ($871) implied by payer coverage for probe testing. Following a negative CME, the NAAT scenario resulted in symptom resolution in more patients (650 per 1000 patients tested) than did either CME (525) or the CME probe testing–based scenario (602) at incremental cost-effectiveness ratios lower than the willingness to pay implied by coverage for CME. Therefore, NAAT is likely to cost-effectively improve health outcomes for patients with vaginitis.

Type 2 diabetes (T2D) prevalence increases unabated even as interventions focused on individuals at high risk can prevent T2D (1). Best approaches to identify individuals at high risk continue to be refined. Current risk assessment focuses on elevated glycemia, commonly estimated by fasting glucose (FG) and HbA1c levels, despite imperfect sensitivity and specificity (2). We previously showed that risk for T2D increases progressively as HbA1c increases, independently and in addition to increasing FG (3,4). As risk assessment for T2D depends on more than glycemia alone, simultaneous consideration of another physiological axis may improve biomarker-based diagnostic precision. Here, we test the hypothesis that simultaneous consideration of HbA1c and insulin resistance (IR), assessed with fasting insulin as HOMA of IR (HOMA-IR), can substantially improve risk assessment for T2D. We have previously detailed our statistical approach, IR and T2D diagnostic criteria, and particulars of the Framingham Heart Study (FHS) (3). Using the same FHS data and excluding those with T2D at baseline, we categorized individuals according to HbA1c <5.7% or 5.7–6.49% and into HOMA-IR tertiles and followed them for a mean (SD) of 16.4 (4.5) years for incident T2D. Age- and sex-adjusted T2D incidence rates and counts of subjects are shown in Fig. 1 A for all …

Because chronic kidney disease (CKD) is underdiagnosed, many patients do not receive care that could slow or prevent progression. Potential CKD patients can be identified during employee wellness events and referred into care by a CKD outreach program. This study assessed the health and economic benefits associated with a CKD outreach program. A model-based cost-effectiveness analysis was conducted for a cohort of patients at risk for CKD under 2 scenarios: wellness events with a CKD outreach program and wellness events without outreach. The outreach program identified potential CKD patients based on estimated glomerular filtration rates. Health outcomes and total cost to payers were estimated with Markov models using 1-year cycles. Because outreach could be offered to either patients with diabetes or to all potential CKD patients, these groups were modeled separately. The authors assumed 40% percent of potential CKD patients accepted the invitation to participate in the CKD outreach program. Model parameters were taken from peer-reviewed literature. The study was conducted from the perspective of self-insured employers over a 5-year time horizon. The study found that the CKD outreach program resulted in a gain of 2.3 quality-adjusted life-years and saved $500,211 when 1000 potential CKD patients with diabetes were invited. When potential CKD patients were invited without regard for diabetes status, 0.8 quality-adjusted life-years were gained at a cost savings of $34,161. The authors concluded that CKD outreach programs can improve health outcomes for patients with CKD and save costs for payers.

Background The American Heart Association and American College of Cardiology guidelines defined patient‐management groups that would benefit from lowering of low‐density lipoprotein cholesterol (LDL‐C). We assessed gaps in dyslipidemia care among employees and spouses with health benefits. Methods and Results We studied 17 889 employees and spouses who were covered by an employer‐sponsored health plan and participated in an annual health assessment. Using medical claims, laboratory tests, and risk assessment questionnaires, we found that 43% of participants were in one of 4 patient‐management groups: secondary prevention, severe hypercholesterolemia (LDL‐C ≥190 mg/dL at least once in the preceding 5 years), diabetes mellitus, or elevated 10‐year risk of cardiovascular disease. To assess gaps in dyslipidemia care, we used LDL‐C ≤70 mg/dL as the goal for both the secondary prevention group and those in the elevated 10‐year risk group with >20% risk; LDL‐C ≤100 mg/dL was used for the other groups. Among those in patient‐management groups, 27.3% were in the secondary prevention group, 7.4% were in the severe hypercholesterolemia group, 29.9% were in the diabetes mellitus group, and 35.4% were in the elevated 10‐year risk group. About 74% of those in patient‐management groups had above‐goal LDL‐C levels, whereas only 31% had evidence of a lipid‐lowering therapy in the past 6 months: 45% in the secondary prevention group, 31% in the severe hypercholesterolemia group, 36% in the diabetes mellitus group, and 17% in the elevated 10‐year risk group. Conclusions The substantial gaps in LDL‐C treatment and goal attainment among members of an employer‐sponsored medical plan who were mostly aware of their LDL‐C levels indicate the need for gap‐closure initiatives.

Background: T2D risk assessment is typically based on fasting glucose (FG) and HbA1c levels. We tested the hypothesis that HbA1c-based T2D risk assessment is improved by also considering another physiological axis, insulin resistance (IR), assessed using a fasting insulin (FI) homeostasis model (HOMA-IR). Methods: We followed 2,205 white individuals from the Framingham Heart Study without T2D (medication use or FG >=126 mg/dL) at baseline for a mean of 16 years. We used age-sex adjusted regression to model incident T2D as a function of baseline HbA1c and IR tertile of HOMA-IR (FI x FG/22.5). Results: Of 266 incident T2D cases, 205 (77%) were in HOMA-IR tertile 3, and of these, 135 (66% of 205) had HbA1c <5.7 (Table). Individuals in HOMA-IR tertile 3 had 16-year T2D CI rates of 20-35%. Both HbA1c and IR were independently associated with elevated incident T2D risk. We saw similar patterns in the 1,583 individuals with baseline FG <100 mg/dL. Conclusion: HbA1c and fasting insulin are commonly available clinical diagnostic tests. Combined consideration of high levels of both identifies the great majority of white individuals at highly increased risk for future T2D. Disclosure J.B. Meigs: Consultant; Self; Quest Diagnostics. B.C. Porneala: None. A. Leong: None. D. Shiffman: Employee; Self; Quest Diagnostics. J. Devlin: Employee; Self; Quest Diagnostics. M.J. McPhaul: Employee; Self; Quest Diagnostics.

Abstract Introduction Obstructive sleep apnea (OSA) is common in individuals with metabolic syndrome (MetS) and increases risk of cardiovascular (CVD) events. Once recognized, therapeutic interventions can reduce OSA severity and associated CVD risk. Of the 25 million Americans with OSA, 80% are unaware of their disease. To facilitate and improve diagnosis of OSA, diagnostic devices for at-home OSA testing have been developed in clinical studies and approved by FDA. We evaluated an employer-sponsored healthcare outreach program and at-home OSA testing as a means of identifying individuals likely to have OSA and referring them into care. Methods Nine-hundred individuals with MetS, positive OSA Berlin questionnaire score and no prior diagnosis of OSA, as determined by annual workplace screening and health claims, were invited to participate in the sleep program. Those who agreed to participate (9.9%) received a diagnostic device for at-home OSA testing. Apnea-hypoapnea index (AHI) results recorded on returned diagnostic devices were evaluated by a sleep specialist. A telephone consultation with a program physician then provided each participant with an explanation of test results and referral into care. Based on AHI we identified individuals with moderate (AHI 16-30) to severe (AHI >30) OSA and referred them to care. Results Of the 89 participating individuals, 21% had 3 MetS components, 53% had 4 components, and 20% had 5 components; 30% were diabetic; 83% had hypertension; and >50% were obese. Moderate to severe OSA was diagnosed in 52 (58%) of participants. Of those, 50% had moderate OSA and 50%, had severe OSA. Among individuals with moderate to severe OSA, 29 (56%) had a physician consultation and were referred to treatment. Conclusion A personalized employer-sponsored healthcare outreach program identified individuals with unrecognized OSA and referred them into care. Support

Introduction: Cardiovascular health (CVH) may be concordant in spouse-pairs because of similar environmental exposures and assortative mating. We investigated the concordance of AHA-defined Life’s ...

Patients with chronic kidney disease (CKD) are often unaware of their condition and therefore do not receive care that could slow or prevent progression. Annual employee wellness events can incorporate a CKD outreach program to help refer those at risk of CKD for appropriate care. This study assessed the potential health and economic benefits of adding a CKD outreach program to employee wellness events. Based on estimated glomerular filtration rate (eGFR) results from annual wellness testing, individuals with an eGFR of <60 mL/min/1.73 m2 were invited to participate in the CKD program. Those who had an eGFR of <60 at two consecutive annual wellness events were referred to care, which was expected to include guideline-supported ACE inhibitors. From a payer perspective, we simulated 5-year health outcomes and cost with Markov models for diabetic CKD patients and non-diabetic CKD patients under 2 scenarios: 1 with the CKD program and 1 without it. Model parameters were based on the actual program or taken from peer-reviewed literature and governmental fee schedules. In the base case 50 percent of those invited accepted referral to care. For 1,000 diabetic CKD patients invited to the CKD program, the program would improve health outcomes with a gain of 0.32 quality-adjusted life-years and save $413,752 dollars. Seven cases of end-stage renal disease (ESRD) would be avoided by year 5. If those with CKD were invited without regard to diabetes status, the CKD program would improve health outcomes at an incremental cost-effectiveness ratio of $33,908 per quality-adjusted life-year gained. Beyond 5 years, opportunities to delay ESRD events would increase, thereby improving the potential CKD program cost-savings for all patients. CKD outreach programs associated with annual wellness events can improve health outcomes for individuals with CKD and could be cost-savings for payers.

Abstract Context Insulin resistance (IR) can progress to type 2 diabetes. Therefore, timely identification of IR could facilitate disease prevention efforts. However, direct measurement of IR is not feasible in a clinical setting. Objective Develop a clinically practical probability score to assess IR in apparently healthy individuals based on levels of insulin, C-peptide, and other risk factors. Design Cross-sectional study. Participants Apparently healthy individuals who volunteered to participate in studies of IR. Main Outcome Measure IR, defined as the top tertile of steady-state plasma glucose during an insulin-suppression test. Results In a study of 535 participants, insulin, C-peptide, creatinine, body mass index (BMI), and triglycerides to high-density lipoprotein cholesterol ratio (TG/HDL-C) were independently associated with IR (all P < 0.05) in a model that included age, sex, ethnicity, BMI, blood pressure, insulin, C-peptide, fasting glucose, low-density lipoprotein cholesterol, TG/HDL-C, alanine aminotransferase, and creatinine. For an IR probability score based on a model that included insulin, C-peptide, creatinine, TG/HDL-C, and BMI, the odds ratio was 26.7 (95% CI 14.0 to 50.8) for those with scores >66% compared with those with scores <33%. When only insulin and C-peptide were included in the model, the odds ratio was 15.6 (95% CI 7.5 to 32.4) for those with scores >66% compared with those with scores <33%. Conclusions An IR probability score based on insulin, C-peptide, creatinine, TG/HDL-C, and BMI or a score based on only insulin and C-peptide may help assess IR in apparently healthy individuals.

Assessment of hemoglobin A1c (HbA1c) levels in addition to fasting glucose (FG) levels and other risk factors can improve diabetes risk assessment (1). Employee wellness programs (EWPs) are common in the U.S. (2), providing opportunities to identify working-age individuals at risk for diabetes (3) and to offer risk-reduction programs targeted to those at most risk. Based on the records of one large laboratory testing provider for EWPs, 25% of those offered FG testing were also offered HbA1c testing (4). Here we investigated whether the addition of HbA1c to FG testing for EWP participants with apparently normal FG (<100 mg/dL) would identify those at elevated risk for incident diabetes. The analysis was based on a cohort of 34,676 employees and spouses who participated in an EWP in 2012. Those with baseline FG ≥100 mg/dL, HbA1c ≥6.5% (48 mmol/mol), or a self-reported physician diagnosis of diabetes ( n = 8,837), with missing baseline data ( n = 244), or who failed to participate in the EWP at least once during 4 years of follow-up ( n = 4,256) were excluded, leaving 21,339 participants. The association …

BACKGROUND:Genetic diagnosis of unexplained global developmental delay and intellectual disability (GDD/ID) often ends the diagnostic odyssey and can lead to changes in clinical management.OBJECTIVE:The objective of this study was to investigate the cost effectiveness of testing scenarios involving several methods used to diagnose GDD/ID: karyotyping, chromosomal microarray analysis (CMA), and targeted next-generation sequencing (NGS).METHODS:We used decision-tree models to estimate the number of genetic diagnoses, the cost from a payers' perspective in the USA, and the incremental cost per additional genetic diagnosis. Model parameters were taken from peer-reviewed literature and governmental fee schedules.RESULTS:CMA testing results in more genetic diagnoses at an incremental cost of US $2692 per additional diagnosis compared with karyotyping, which has an average cost per diagnosis of US $11,033. Performing both tests sequentially results in the same number of diagnoses, but the total cost is less when CMA testing is done first and karyotyping second. Furthermore, when CMA testing yields a variant of unknown significance, additional genetic diagnoses can be obtained at an incremental cost of US $4220 by CMA testing of both parents, and when parents are not available or the patient had a normal CMA result, targeted NGS of the patient can add diagnoses at a further incremental cost of US $12,295.CONCLUSION:These results provide a cost effectiveness rationale for the use of CMA as the first-tier test for the genetic diagnosis of unexplained GDD/ID and further indicate that testing of both parents may be cost effective when a variant of unknown significance is detected in the patient.

OBJECTIVE:Hemoglobin A1c (HbA1c) can be used to assess type 2 diabetes (T2D) risk. We asked whether HbA1c was associated with T2D risk in four scenarios of clinical information availability: 1) HbA1c alone, 2) fasting laboratory tests, 3) clinic data, and 4) fasting laboratory tests and clinic data. RESEARCH DESIGN AND METHODS:We studied a prospective cohort of white (N = 11,244) and black (N = 2,294) middle-aged participants without diabetes in the Framingham Heart Study and Atherosclerosis Risk in Communities study. Association of HbA1c with incident T2D (defined by medication use or fasting glucose [FG] ≥126 mg/dL) was evaluated in regression models adjusted for 1) age and sex (demographics); 2) demographics, FG, HDL, and triglycerides; 3) demographics, BMI, blood pressure, and T2D family history; or 4) all preceding covariates. We combined results from cohort and race analyses by random-effects meta-analyses. Subsidiary analyses tested the association of HbA1c with developing T2D within 8 years or only after 8 years. RESULTS:Over 20 years, 3,315 individuals developed T2D. With adjustment for demographics, the odds of T2D increased fourfold for each percentage-unit increase in HbA1c. The odds ratio (OR) was 4.00 (95% CI 3.14, 5.10) for blacks and 4.73 (3.10, 7.21) for whites, resulting in a combined OR of 4.50 (3.35, 6.03). After adjustment for fasting laboratory tests and clinic data, the combined OR was 2.68 (2.15, 3.34) over 20 years, 5.79 (2.51, 13.36) within 8 years, and 2.23 (1.94, 2.57) after 8 years. CONCLUSIONS:HbA1c predicts T2D in different common scenarios and is useful for identifying individuals with elevated T2D risk in both the short- and long-term.

Genomic information has the potential to have an impact on many aspects of health care, but the marked variation present in the human genome, corresponding to a variant found in every eight nucleotides sequenced [[1]Lek M. Karczewski K.J. Minikel E.V. et al.Analysis of protein-coding genetic variation in 60,706 humans.Nature. 2016; 536: 285-291Crossref PubMed Scopus (6551) Google Scholar], highlights the complexity and importance of context surrounding economic evaluation of genome-based technologies. The recent article by Li et al. [[2]Li Y. Arellano A.R. Bare L.A. et al.A multigene test could cost-effectively help extend life expectancy for women at risk of hereditary breast cancer.Value Health. 2017; 20: 547-555Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar] demonstrated an incremental cost-effectiveness ratio of $42,067 per life-year gained by testing unaffected women with a family history of breast cancer (BC) with a seven-gene familial BC panel including the BRCA1 and BRCA2 genes, compared with BRCA1 and BRCA2 testing only. Here, we describe our significant concerns with the parameter inputs used by Li et al. in their economic evaluation and highlight heterogeneity and uncertainty specific to genome-based evaluations that need to be considered, especially when combining data from multiple populations [3LaDuca H. Stuenkel A.J. Dolinsky J.S. et al.Utilization of multigene panels in hereditary cancer predisposition testing: analysis of more than 2,000 patients.Genet Med. 2014; 16: 830-837Abstract Full Text Full Text PDF PubMed Scopus (251) Google Scholar, 4Hall M.J. Reid J.E. Burbidge L.A. et al.BRCA1 and BRCA2 mutations in women of different ethnicities undergoing testing for hereditary breast-ovarian cancer.Cancer. 2009; 115: 2222-2233Crossref PubMed Scopus (272) Google Scholar]. The areas that markedly alter the accuracy and general applicability of this model are 1) variant classification uncertainty, 2) population heterogeneity, and 3) genotype-specific biological relevance. The classification of genetic variants as either pathogenic or nonpathogenic can be complex. The first standard variant classification system was proposed in 2008 [[5]Plon S.E. Eccles D.M. Easton D. et al.Sequence variant classification and reporting: recommendations for improving the interpretation of cancer susceptibility genetic test results.Hum Mutat. 2008; 29: 1282-1291Crossref PubMed Scopus (659) Google Scholar], leading to the development of clinical guidelines that have been adopted in many countries. This has resulted in a more consistent, transparent international approach to classification [6Richards C.S. Bale S. Bellissimo D.B. et al.ACMG recommendations for standards for interpretation and reporting of sequence variations: revisions 2007.Genet Med. 2008; 10: 294-300Abstract Full Text Full Text PDF PubMed Scopus (633) Google Scholar, 7Richards S. Aziz N. Bale S. et al.Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology.Genet Med. 2015; 17: 405-424Abstract Full Text Full Text PDF PubMed Scopus (14601) Google Scholar]. This system integrates parameters relating to how the variant affects protein function, segregates with disease within families, is conserved through evolution, and is novel or present at very low frequencies in population variant databases. This latter filter means that variants found in ethnic groups not well represented in such data sets [[1]Lek M. Karczewski K.J. Minikel E.V. et al.Analysis of protein-coding genetic variation in 60,706 humans.Nature. 2016; 536: 285-291Crossref PubMed Scopus (6551) Google Scholar] are often associated with increased classification uncertainty and possible higher false pathogenic classification rates. In the model by Li et al., the populations used in the standard testing (Hall et al. [[4]Hall M.J. Reid J.E. Burbidge L.A. et al.BRCA1 and BRCA2 mutations in women of different ethnicities undergoing testing for hereditary breast-ovarian cancer.Cancer. 2009; 115: 2222-2233Crossref PubMed Scopus (272) Google Scholar]) and comparator (LaDuca et al. [[3]LaDuca H. Stuenkel A.J. Dolinsky J.S. et al.Utilization of multigene panels in hereditary cancer predisposition testing: analysis of more than 2,000 patients.Genet Med. 2014; 16: 830-837Abstract Full Text Full Text PDF PubMed Scopus (251) Google Scholar]) arms varied significantly in ethnic composition and method of variant classification (Table 1). BRCA1 and BRCA2 mutation prevalence data from Hall et al. predated the now-accepted classification system, and women from ethnic groups not well represented in variant databases, African and Latin American, demonstrated the highest BRCA1 mutation prevalence (10.2% and 9.6%, respectively) compared with the white subgroup (7.3%). In contrast, LaDuca et al. directly referenced their adoption of the international variant classification guidelines.Table 1Proportion of high-risk parameters in the populations from which mutation prevalence was derived for modelPopulation parametersStandard testing arm [4]Hall M.J. Reid J.E. Burbidge L.A. et al.BRCA1 and BRCA2 mutations in women of different ethnicities undergoing testing for hereditary breast-ovarian cancer.Cancer. 2009; 115: 2222-2233Crossref PubMed Scopus (272) Google Scholar (%)Panel testing arm [3]LaDuca H. Stuenkel A.J. Dolinsky J.S. et al.Utilization of multigene panels in hereditary cancer predisposition testing: analysis of more than 2,000 patients.Genet Med. 2014; 16: 830-837Abstract Full Text Full Text PDF PubMed Scopus (251) Google Scholar (%)Ethnicity White87.173.1 Ashkenazi Jewish06.8 Other12.9Cancer-affected All8695.1 Breast cancer6192.7 Ovarian cancer71 Breast and ovarian cancer2.40Age at diagnosis (y) <304.1 31–4016.4 41–5027.3 <5040 (breast cancer only)47.8 (all cases) >5020 (breast cancer only)52.3 (all cases) Open table in a new tab An additional confounder is that the molecular laboratory in the study by Hall et al. used an undisclosed proprietary variant classification process. This nontransparency was further compounded by the absence of information on the pathogenic variants in the original article and the lack of contribution by this laboratory to open variant databases such as ClinVar [[8]Landrum M.J. Lee J.M. Riley G.R. et al.ClinVar: public archive of relationships among sequence variation and human phenotype.Nucleic Acids Res. 2014; 42: D980-D985Crossref PubMed Scopus (1628) Google Scholar]. This means that variant pathogenicity could not be independently assessed and verified, adding significant uncertainty to the predicted/estimated mutation prevalence cited. The prevalence of germline BC predisposition gene mutations is highly dependent on 1) population parameters such as the proportion of cancer-affected individuals, the cancer family history of individuals, and the ethnic mix of the population and 2) cancer-specific parameters including the organ-specific type of cancer (breast vs. ovarian cancer), the age at diagnosis, and cancer histopathology (Table 2) [9Tung N. Lin N.U. Kidd J. et al.Frequency of germline mutations in 25 cancer susceptibility genes in a sequential series of patients with breast cancer.J Clin Oncol. 2016; 34: 1460-1468Crossref PubMed Scopus (321) Google Scholar, 10Song H. Cicek M.S. Dicks E. et al.The contribution of deleterious germline mutations in BRCA1, BRCA2 and the mismatch repair genes to ovarian cancer in the population.Hum Mol Genet. 2014; 23: 4703-4709Crossref PubMed Scopus (97) Google Scholar, 11Wong-Brown M.W. Meldrum C.J. Carpenter J.E. et al.Prevalence of BRCA1 and BRCA2 germline mutations in patients with triple-negative breast cancer.Breast Cancer Res Treat. 2015; 150: 71-80Crossref PubMed Scopus (91) Google Scholar, 12Haffty B.G. Choi D.H. Goyal S. et al.Breast cancer in young women (YBC): prevalence of BRCA1/2 mutations and risk of secondary malignancies across diverse racial groups.Ann Oncol. 2009; 20: 1653-1659Crossref PubMed Scopus (47) Google Scholar, 13Manchanda R. Loggenberg K. Sanderson S. et al.Population testing for cancer predisposing BRCA1/BRCA2 mutations in the Ashkenazi-Jewish community: a randomized controlled trial.J Natl Cancer Inst. 2015; 107: 379Crossref PubMed Scopus (137) Google Scholar, 14McClain M.R. Palomaki G.E. Nathanson K.L. et al.Adjusting the estimated proportion of breast cancer cases associated with BRCA1 and BRCA2 mutations: public health implications.Genet Med. 2005; 7: 28-33Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar]. In Li et al., the mutation prevalence data came from two discrete populations that differed significantly not only in ethnic composition but also in age and personal cancer history and type.Table 2BRCA1 and BRCA2 mutation prevalence in unselected cancer cases (with or without a family history)PopulationClinical featuresBRCA1/2 mutation prevalence (%)ReferenceBreast cancerAll invasive cases6.1[5]Plon S.E. Eccles D.M. Easton D. et al.Sequence variant classification and reporting: recommendations for improving the interpretation of cancer susceptibility genetic test results.Hum Mutat. 2008; 29: 1282-1291Crossref PubMed Scopus (659) Google ScholarYoung (diagnosed at <45 y)13.0[8]Landrum M.J. Lee J.M. Riley G.R. et al.ClinVar: public archive of relationships among sequence variation and human phenotype.Nucleic Acids Res. 2014; 42: D980-D985Crossref PubMed Scopus (1628) Google ScholarTriple-negative breast cancer9.6[7]Richards S. Aziz N. Bale S. et al.Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology.Genet Med. 2015; 17: 405-424Abstract Full Text Full Text PDF PubMed Scopus (14601) Google ScholarAshkenazi Jewish18.4[5]Plon S.E. Eccles D.M. Easton D. et al.Sequence variant classification and reporting: recommendations for improving the interpretation of cancer susceptibility genetic test results.Hum Mutat. 2008; 29: 1282-1291Crossref PubMed Scopus (659) Google ScholarOvarian cancerAll epithelial cancer cases8.4[6]Richards C.S. Bale S. Bellissimo D.B. et al.ACMG recommendations for standards for interpretation and reporting of sequence variations: revisions 2007.Genet Med. 2008; 10: 294-300Abstract Full Text Full Text PDF PubMed Scopus (633) Google ScholarHigh-grade serous12.4[6]Richards C.S. Bale S. Bellissimo D.B. et al.ACMG recommendations for standards for interpretation and reporting of sequence variations: revisions 2007.Genet Med. 2008; 10: 294-300Abstract Full Text Full Text PDF PubMed Scopus (633) Google ScholarUnaffectedAshkenazi Jewish2.45[9]Tung N. Lin N.U. Kidd J. et al.Frequency of germline mutations in 25 cancer susceptibility genes in a sequential series of patients with breast cancer.J Clin Oncol. 2016; 34: 1460-1468Crossref PubMed Scopus (321) Google ScholarWhite population0.22–0.37[14]McClain M.R. Palomaki G.E. Nathanson K.L. et al.Adjusting the estimated proportion of breast cancer cases associated with BRCA1 and BRCA2 mutations: public health implications.Genet Med. 2005; 7: 28-33Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar Open table in a new tab The input data for the baseline BRCA1/2-only genetic testing arm were derived from a large population referred for testing between 1996 and 2006, of which 87.1% were white. The proportion of women of Ashkenazi Jewish ethnicity in a population can significantly alter BRCA mutation prevalence. As a high-risk ethnic group, a cancer-naive Ashkenazi woman has a 2% risk of carrying one of three BRCA founder mutations, increasing to 45% if she has ovarian cancer and a family cancer history [[15]Levy-Lahad E. Catane R. Eisenberg S. et al.Founder BRCA1 and BRCA2 mutations in Ashkenazi Jews in Israel: frequency and differential penetrance in ovarian cancer and in breast-ovarian cancer families.Am J Hum Genet. 1997; 60: 1059-1067PubMed Google Scholar]. The corollary of this is that once these mutations are excluded, the likelihood of detecting an alternative high-risk gene mutation is low. Although Hall et al. excluded this ethnic group from the BRCA-only testing group, this was based on the ethnicity reported on test requisition forms, which is not a reliable means for ascertaining Ashkenazi Jewish ancestry and is likely to lead to misclassification. This concern would appear valid because 12.6% of BRCA mutations in the white subgroup were Ashkenazi Jewish founder mutations. Equally contentiously, this ethnic group comprised 7% of the BRCA mutation negative comparator population. The model developed by Li et al. related to testing an unaffected female population aged 40 or 50 years, yet the model input parameters were obtained from cohorts in which most had a personal cancer history. The population in the study by Hall et al. contained 86% with a personal diagnosis of cancer: 81.3% BC, 4.7% ovarian cancer, and only 14% being unaffected. On average, 47% of women with BC were diagnosed when they were younger than 50 years. In contrast, mutation data for the seven-gene panel testing comparator were taken from a population of 874 individuals (97.1% women), previously shown not to harbor a BRCA mutation, who were tested between 2012 and 2013. The difference in time periods over which these two populations were tested means that changes in clinical BRCA testing guidelines after 2006, such as the recent emphasis on testing triple-negative BC (triple negative tumour (TNT): hormone receptor negative and HER2 unamplified histopathology), will have altered the underlying characteristics of this population. In this group, 95.1% had a personal diagnosis of cancer, 92.7% BC (9.3% of these being triple- negative cancers), and 1% ovarian cancer. A fifth of the women were diagnosed when they were younger than 40 years. In both populations, a significant proportion (20.5%–47.8%) of women were both affected and tested at a younger age than the target baseline age for the unaffected population modeled by Li et al. These data should have been excluded from use in mutation prevalence input estimates on these grounds. In addition, because the model developed by Li et al. relates to testing unaffected individuals, the lower mutation prevalence value used in the sensitivity analysis (1.2%) may be a more conservative base-case estimate, because it would align with the carrier probability of an unaffected first-degree relative of an affected mutation-positive individual. There are indications that mutation prevalence may be even lower, as found by Desmond et al. [[16]Desmond A. Kurian A.W. Gabree M. et al.Clinical actionability of multigene panel testing for hereditary breast and ovarian cancer risk assessment.JAMA Oncol. 2015; 1: 943-951Crossref PubMed Scopus (249) Google Scholar], where only 0.76% (8 of 1046) of non-BRCA1/2 carriers harbored a mutation in any one of the five genes analyzed, although again this study included only 14% of unaffected individuals. The impact of each factor on the incremental diagnostic yield produced by panel testing compared with BRCA1/2-only testing is difficult to assess because of the noncomparable nature of the populations used in the analysis. We accept that because high-risk mutation carriers are rare in the population, models incorporating data from disparate populations may be required. In such circumstances, we suggest that care be taken to standardize the model inputs from different populations by weighting the mutation prevalence by the proportion of known high-risk factors within each population. In addition to cancer risk, germline BC mutations often determine, both directly and indirectly, cancer prognosis. In this model, baseline BC mortality was calculated from the Surveillance, Epidemiology, and End Results Program data, and did not adjust for adverse prognostic features common in BRCA1/2-associated cancers such as high grade and younger age of onset, and higher likelihood of TNT for BRCA1 [[17]Brekelmans C.T. Tilanus-Linthorst M.M. Seynaeve C. et al.Tumour characteristics, survival and prognostic factors of hereditary breast cancer from BRCA2-, BRCA1- and non-BRCA1/2 families as compared to sporadic breast cancer cases.Eur J Cancer. 2007; 43: 867-876Abstract Full Text Full Text PDF PubMed Scopus (183) Google Scholar]. Women are at increased risk of cancer-specific death for at least 10 years after a breast or ovarian cancer diagnosis. The model accounted for this by assuming that a patient returned to the well state if alive 5 years postdiagnosis, with a 5% recurrence rate in the next 5 years. This may have again underestimated cancer-specific mortality, because this recurrence rate was based on BC cases in an older population, who are likely to have lower grade, less aggressive cancers than mutation carriers. It was also less relevant for women diagnosed with ovarian cancer, where it is estimated that an additional 13% of patients would succumb to their disease between 5 and 10 years after diagnosis [[18]Baldwin L.A. Huang B. Miller R.W. et al.Ten-year relative survival for epithelial ovarian cancer.Obstet Gynecol. 2012; 120: 612-618Crossref PubMed Scopus (221) Google Scholar]. The model also did not account for competing gene-specific, non-BC mortality rates. Any difference in life-years gained resulting from BC risk management may be negated in mutation carriers in high-risk genes CDH1 and TP53 because of the competing increased susceptibility to cancers with high mortality, such as diffuse gastric cancer and sarcoma, respectively. Given the wide spectrum of cancers across the hereditary syndromes included in the analysis, it was reasonable to exclude direct modeling of additional cancers, but it is necessary to account for their impact on life expectancy to provide a more accurate clinical picture. Limiting the model scope to BC mortality only was not explicitly justified, and it would be of interest to assess the impact of a more realistic mortality model on the incremental cost-effectiveness ratio obtained. Genome-based technologies have the potential to have a significant impact on many aspects of clinical care and disease prevention, but their often probabilistic outputs need specialist care in modeling and evaluating in a context-dependent manner.