Inpatient COVID-19 Mortality Rates: What are the predictors?

ObjectiveThis study aims to investigate the relationship between registered nurses and hospital-based medical specialties staffing levels with inpatient COVID-19 mortality rates. MethodsWe rely on data from AHA Annual Survey Database, Area Health Resource File, and UnitedHealth Group Clinical Discovery Database. We use linear regression to analyze the association between hospital staffing levels and bed capacity with inpatient COVID-19 mortality rates from March 1, 2020, through December 31, 2020. ResultsHigher staffing levels of registered nurses, hospitalists, and emergency medicine physicians were associated with lower COVID-19 mortality rates. Moreover, a higher number of ICU and skilled nursing beds were associated with better patient outcomes. Hospitals located in urban counties with high infection rates had the worst patient mortality rates. ConclusionHigher staffing levels are associated with lower inpatient mortality rates for COVID-19 patients. A future assessment is needed to establish benchmarks on the minimum staffing levels for nursing and hospital-based medical specialties during pandemics.

dPQL: a lossless distributed algorithm for generalized linear mixed model with application to privacy-preserving hospital profiling

Hospital profiling provides a quantitative comparison of health care providers for their quality of care regarding certain clinical outcomes. To implement hospital profiling, the generalized linear mixed model (GLMM) is usually used to fit clinical or administrative claims data, adjusting for the effects of covariates. For better generalizability, data across multiple hospitals, databases or networks are desired. However, due to the privacy regulation and the computation complexity of GLMM, a convenient distributed algorithm for hospital profiling is needed. In this paper, we develop a novel distributed Penalized Quasi Likelihood algorithm (dPQL) to fit GLMM, when only aggregated data, rather than the individual patient data are available across hospitals. The dPQL algorithm is based on a newly-developed distributed linear mixed model (DLMM) algorithm. This proposed dPQL algorithm is lossless, i.e. it obtains identical results as if the individual patient data are pooled from all hospitals. We demonstrate the usage of the dPQL algorithms by ranking 929 hospitals for COVID-19 mortality or referral to hospice in Asch, et al. 2020.

Patient Characteristics in Cases of Reinfection or Prolonged viral shedding in SARS-CoV-2

ImportanceAs testing options increase for COVID-19, their interpretability is challenged by the increasing variety of clinical contexts in which results are obtained. In particular, positive COVID-19 diagnostic (RT-PCR) tests that occur after a patient has seroconverted may be indicative of reinfection. However, in the absence of SARS-CoV-2 sequence data, the possibility of prolonged viral shedding may not be excluded. We highlight a testing pattern that identifies such cases and study its statistical power in identifying potential reinfection. We also study the medical records of patients that matched the pattern. ObjectiveTo describe the frequency and demographic information of people with a testing pattern indicative of SARS-CoV-2 reinfection. DesignWe examined 4.2 million test results from a large national health insurer in the United States. Specifically, we identified the pattern of a positive RT-PCR test followed by a positive IgG test, again followed by a positive RT-PCR. SettingData from outpatient laboratories across the United States was joined with claims data from a single large commercial insurers administrative claims database. ParticipantsStudy participants are those whose insurance, either commercial or Medicare, is provided by a single US based insurer. ExposuresPeople who received at least two positive diagnostic tests via RT-PCR for SARS-Cov-2 separated by 42 or more days with at least one serological test (IgG) indicating the presence of antibodies between diagnostic tests. Main Outcomes and MeasuresCount and characteristics of people with the timeline of three tests as described in Exposures. ResultsWe identified 79 patients who had two positive RT-PCR tests separated by more than six weeks, with a positive IgG test in between. These patients tended to be older than those COVID-19 patients without this pattern (median age 56 vs. 42), and they exhibited comorbidities typically attributed to a compromised immune system and heart disease. Conclusions and RelevanceWhile the testing pattern alone was not sufficient to distinguish potential reinfection from prolonged viral shedding, we were able to identify common traits of the patients identified through the pattern.

Pooling-in-a-pod: A Strategy for COVID-19 Testing to Facilitate Safe Return to School

The COVID-19 pandemic has prompted widespread primary and secondary school closures. Routine testing of asymptomatic students and staff, as part of a comprehensive program, can help schools open safely. "Pooling-in-a-pod" is a public health surveillance strategy whereby testing cohorts are composed based on social relationships and physical proximity. Pooling-in-a-pod allowed for weekly on-site point-of-care testing of all staff and students at an independent preschool to grade 12 school in Washington, D.C. Staff and older students self-collected anterior nares samples, and trained staff collected samples from younger students. Overall, 6,746 samples were tested for 815 students and 276 staff between November 30, 2020, and March 3, 2021. The average pool size was 7.3 people. Sample collection to pool result time averaged 40 minutes. The direct testing cost per person per week was $24.77, including swabs. One surveillance test pool was positive. During the study period, daily new cases in Washington, D.C., ranged from 24 - 46 per 100,000 population. A post-launch survey found most parents (90.3%), students (93.4%), and staff (98.8%) were willing to participate in pooled testing with confirmatory tests for positive pool members. The school reported a 32.6% decrease in virtual learning after initiation of the program. Pooling-in-a-pod is feasible, cost-effective, and an acceptable COVID-19 surveillance strategy for schools. School officials and policymakers can leverage this strategy to facilitate safe, sustainable, in-person schooling. SUMMARY 1) What is the current understanding of this subject?Routine COVID-19 testing as part of a comprehensive strategy to operate schools safely is currently not widely implemented. 2) What does this report add to the literature?"Pooling-in-a-pod," is a public health surveillance strategy whereby cohorts are composed based on social relationships and physical proximity. 6,746 samples were tested in 969 pools (average pool size 7.3 people) in a Washington, D.C. school, thereby requiring fewer test kits and less expense. The program was widely acceptable. 3) What are the implications for public health practice?Pooling-in-a-pod allows for more accessible testing to facilitate safe in-person schooling and minimize the negative effects of distance learning.

Identifying Optimal COVID-19 Testing Strategies for Schools and Businesses: Balancing Testing Frequency, Individual Test Technology, and Cost

BackgroundCOVID-19 test sensitivity and specificity have been widely examined and discussed yet optimal use of these tests will depend on the goals of testing, the population or setting, and the anticipated underlying disease prevalence. We model various combinations of key variables to identify and compare a range of effective and practical surveillance strategies for schools and businesses. MethodsWe coupled a simulated data set incorporating actual community prevalence and test performance characteristics to a susceptible, infectious, removed (SIR) compartmental model, modeling the impact of base and tunable variables including test sensitivity, testing frequency, results lag, sample pooling, disease prevalence, externally-acquired infections, and test cost on outcomes case reduction. ResultsIncreasing testing frequency was associated with a non-linear positive effect on cases averted over 100 days. While precise reductions in cumulative number of infections depended on community disease prevalence, testing every 3 days versus every 14 days (even with a lower sensitivity test) reduces the disease burden substantially. Pooling provided cost savings and made a high-frequency approach practical; one high-performing strategy, testing every 3 days, yielded per person per day costs as low as $1.32. ConclusionsA range of practically viable testing strategies emerged for schools and businesses. Key characteristics of these strategies include high frequency testing with a moderate or high sensitivity test and minimal results delay. Sample pooling allowed for operational efficiency and cost savings with minimal loss of model performance.

Variation across population subgroups of COVID-19 antibody testing performance

Understanding variations in the performance of serological tests for SARS-CoV-2 across varying demographics is relevant to clinical interpretations and public policy derived from their results. Appropriate use of serological assays to detect anti-SARS-CoV-2 antibodies requires estimation of their accuracy over large populations and an understanding of the variance in performance over time and across demographic groups. In this manuscript we focus on anti-SARS-CoV-2 IgG, IgA, and IgM antibody tests approved under emergency use authorizations and determine the recall of the serological tests compared to RT-PCR tests by Logical Observation Identifiers Names and Codes (LOINCs). Variability in test performance was further examined over time and by demographics. The recall of the most common IgG assay (LOINC 94563-4) was 91.2% (95% CI: 90.5%, 91.9%). IgA (LOINC 94562-6) and IgM (94564-2) assays performed significantly worse than IgG assays with estimated recall rates of 20.6% and 27.3%, respectively. A statistically significant difference in recall (p = 0.019) was observed across sex with a higher recall in males than females, 92.1% and 90.4%, respectively. Recall also differed significantly by age group, with higher recall in those over 45 compared to those under 45, 92.9% and 88.0%, respectively (p< 0.001). While race was unavailable for the majority of the individuals, a significant difference was observed between recall in White individuals and Black individuals (p = 0.007) and White individuals and Hispanic individuals (p = 0.001). The estimates of recall were 89.3%, 95.9%, and 94.2% for White, Black, and Hispanic individuals respectively.