A mathematical modeling study concluded that when it comes to COVID-19 surveillance, test frequency and turnaround time are considerably more important than test sensitivity (Sci Adv 2020;10.1126/sciadv.abd5393).
The investigators determined that testing 75% of individuals every 3 days using a rapid test with a limit of detection (LOD) of 105 copies/mL and same-day results in an environment with 4% SARS-CoV-2 prevalence and a reproduction rate starting at RO 1.5 would “drive the epidemic toward extinction within 6 weeks and reduce the cumulative incidence by 88%.” In contrast, a similar testing frequency using polymerase chain reaction (PCR) with an LOD of 103 copies/mL and a 48-hour results turnaround would reduce infectiousness by 58%.
“When it comes to public health, it’s better to have a less sensitive test with results today than a more sensitive one with results tomorrow,” said lead author Daniel Larremore, in a prepared statement.
Larremore and his colleagues at the University of Colorado and Harvard University hypothesized that given the features of viral increase, infectivity, and decline during SARS-CoV-2 infection, there would be “minimal differences” in effective screening regimens between PCR tests and cheaper, faster tests with higher LODs.
The investigators modeled viral loads and infectiousness curves for 10,000 simulated individuals based on within-host viral kinetics features like latency and growth. They also assessed the impact of repeated screening at different intervals and with tests of different sensitivities in a university-type setting of 20,000 people and in a large city of 8.4 million. In addition, the researchers modeled the impact on transmission dynamics of factors such as delayed results, changed model assumptions like RO, and the effect of repeated population screening.
Even weekly testing using a test with 100 times lower molecular sensitivity than PCR with just half of a population participating would reduce the peak and length of an outbreak.
Based on their findings, the authors suggested that federal and state governments encourage the development and use of rapid, lower cost, and lower sensitivity tests for public health and repeat population screening.
Wide Regional Differences in Adopting Gene Expression Testing for Prostate Cancer
Amid overall increased use, significant regional variations exist in commercial gene expression testing for prostate cancer, with a greater than 8-fold difference in use between high-adopting and low-adopting regions (JAMA Oncol 2020;doi:10.1001/jamaoncol.2020.6086).
The researchers accessed administrative claims data from Blue Cross Blue Shield Axis for 92,418 men ages 40 to 89 diagnosed with prostate cancer between 2012 to 2018, using CPT codes to identify claims for genomic testing within 6 months of initial diagnosis. During the study period this testing rose from 0.8% of patients to 11.3%.
The researchers analyzed the data according to 217 hospital referral regions (HRR), which reflect regional healthcare markets for tertiary medical care. From this process they identified five distinct regional trajectories of test adoption. Less than 1% of patients were tested in all five groups at baseline, but adoption of this testing increased during the 6-year study period from 4% in the lowest adopting group to 33.8% in the highest.
In comparison to HRRs with the lowest test adoption, those in the highest had higher HRR education measures, household income, and prostate cancer resources, such as density of providers and rates of prostate-specific antigen testing.
The authors did not find a direct link between race and the trajectories but cautioned that this could be due to the study sample of younger, commercially insured patients, which might mitigate racial disparities in accessing cancer care. They also didn’t have access to patient-level demographics, so couldn’t rule-out the possibility of racial differences.
Unrelated geographic regions shared testing adoption trajectories, suggesting similar local level conditions might promote dissemination of new technologies, like access to research-oriented medical centers, relationships with industry, or interest among patients.
Single-page Artificial Pancreas Dashboard Proposed for Reporting Key Glucose Management Metrics
A standardized single-page reporting format for seven glucose metrics across all hybrid closed-loop (HCL) glucose management systems could boost uptake and appropriate use of the systems, ultimately improving patient care, according to the researchers who proposed this report, dubbed Artificial Pancreas Dashboard (AP Dashboard) (Diabetes Technol Ther 2021;23:10.1089/dia.2020.0622). In proposing AP Dashboard, Viral Shah, MD, and Satish Garg, MD, University of Colorado (CU) faculty members and practitioners at the CU Barbara Davis Center for Diabetes, noted that despite advances in therapies and technologies, diabetes outcomes have not improved markedly, and only about one-third of patients achieve optimal glycemic control. While recognizing socioeconomic barriers to accessing newer technologies, the authors also posited that providers’ lack of knowledge could be a major barrier. Finding easier ways to interpret glucose and insulin metrics and optimize HCL settings could help, they suggested.
The seven AP Dashboard components would include: glucose metrics; hypoglycemia; insulin; user experience; hyperglycemia; glucose modal-day profile; and insight. The proposed glucose metrics include: mean glucose; standard deviation (SD) and/or coefficient of variation (CV); glucose management indicator (GMI); and a visual graph of continuous glucose monitor (CGM) metrics like time in range, above range, and below range.
Most members of an expert panel in 2012 agreed that mean glucose is a simple metric that both patients and physicians understand.
While CV is more constant and not affected by mean glucose or HbA1c, the authors believe that patients and clinicians find SD easier to understand. GMI is the new logistic regression formula to estimate HbA1c based on mean glucose.
In a break from current practices, the authors propose including in AP Dashboard both level 1 hypoglycemia (CGM glucose <70 mg/dL) and level 2 hypoglycemia (CGM glucose <54 mg/dL). Just reporting level 1 hypoglycemia doesn’t provide enough information to develop an action plan for minimizing hypoglycemia, according to Shah and Garg. To avoid over-reporting hypoglycemic events they proposed reporting level 1 episodes that last at least 15 minutes.