Association of the Intermountain Risk Score with major adverse health events in patients positive for COVID-19: an observational evaluation of a US cohort

OBJECTIVES

The Intermountain Risk Score (IMRS), composed using published sex-specific weightings of parameters in the complete blood count (CBC) and basic metabolic profile (BMP), is a validated predictor of mortality. We hypothesised that IMRS calculated from prepandemic CBC and BMP predicts COVID-19 outcomes and that IMRS using laboratory results tested at COVID-19 diagnosis is also predictive.

DESIGN

Prospective observational cohort study.

SETTING

Primary, secondary, urgent and emergent care, and drive-through testing locations across Utah and in sections of adjacent US states. Viral RNA testing for SARS-CoV-2 was conducted from 3 March to 2 November 2020.

PARTICIPANTS

Patients aged ≥18 years were evaluated if they had CBC and BMP measured in 2019 and tested positive for COVID-19 in 2020.

PRIMARY AND SECONDARY OUTCOME MEASURES

The primary outcome was a composite of hospitalisation or mortality, with secondary outcomes being hospitalisation and mortality separately.

RESULTS

Among 3883 patients, 8.2% were hospitalised and 1.6% died. Subjects with low, mild, moderate and high-risk IMRS had the composite endpoint in 3.5% (52/1502), 8.6% (108/1256), 15.5% (152/979) and 28.1% (41/146) of patients, respectively. Compared with low-risk, subjects in mild-risk, moderate-risk and high-risk groups had HR=2.33 (95% CI 1.67 to 3.24), HR=4.01 (95% CI 2.93 to 5.50) and HR=8.34 (95% CI 5.54 to 12.57), respectively. Subjects aged <60 years had HR=3.06 (95% CI 2.01 to 4.65) and HR=7.38 (95% CI 3.14 to 17.34) for moderate and high risks versus low risk, respectively; those ≥60 years had HR=1.95 (95% CI 0.99 to 3.86) and HR=3.40 (95% CI 1.63 to 7.07). In multivariable analyses, IMRS was independently predictive and was shown to capture substantial risk variation of comorbidities.

CONCLUSIONS

IMRS, a simple risk score using very basic laboratory results, predicted COVID-19 hospitalisation and mortality. This included important abilities to identify risk in younger adults with few diagnosed comorbidities and to predict risk prior to SARS-CoV-2 infection.

Feasibility of combining serial smartphone single-lead electrocardiograms for the diagnosis of ST-elevation myocardial infarction

BACKGROUND

The rate-limiting step in STEMI diagnosis often is the availability of a 12-lead electrocardiogram (ECG) and its interpretation. The potential may exist to speed the availability of 12-lead ECG information by using commonly available mobile technologies. We sought to test whether combining serial smartphone single-lead ECGs to create a virtual 12-lead ECG can accurately diagnose STEMI.

METHODS

Consenting patients presenting with symptoms consistent with a possible STEMI had contemporaneous standard 12-lead and smartphone '12-lead equivalent' ECG (produced by electronically combining serial single-lead ECGs) recordings obtained. Matched ECGs were evaluated qualitatively and quantitatively by a panel of blinded readers and classified as STEMI/STEMI equivalent (LBBB), Not-STEMI, or uninterpretable. Interpretable ECG pairs were graded as showing good, fair, or poor correlation.

RESULTS

Two hundred four subjects (age = 60 years, males = 57%, STEMI activation = 45%) were enrolled from 5 international sites. Smartphone ECG quality was graded as good in 151 (74.0%), fair in 32 (15.7%), poor in 8 (3.9%), and uninterpretable in 13 (6.4%). A STEMI/STEMI equivalent diagnosis was identified by standard 12-lead ECG in 57/204 (27.9%) recordings. For all interpretable pairs of smartphone ECGs compared with standard ECGs (n = 190), the sensitivity, specificity, and positive and negative predictive values for STEMI/STEMI equivalent by smartphone were 0.89, 0.84, 0.70 and 0.95, respectively.

CONCLUSIONS

A '12-lead equivalent' ECG obtained from multiple serial single-lead ECGs from a smartphone can identify STEMI with good correlation to a standard 12-lead ECG. This technology holds promise to improve outcomes in STEMI by enhancing the reach and speed of diagnosis and thereby early treatment.

Improvement in the predictive ability of the Intermountain Mortality Risk Score by adding routinely collected laboratory tests such as albumin, bilirubin, and white cell differential count

Background:

The Intermountain Mortality Risk Score (IMRS), a sex-specific mortality-prediction metric, has proven to be effective in various populations. IMRS is comprised of the complete blood count (CBC), basic metabolic panel (BMP), and age. Whether the addition of factors from the comprehensive metabolic panel (CMP) and white blood cell (WBC) differential count improves risk stratification is unknown.

Methods:

Patients with baseline complete metabolic panel (CMP) and IMRS measurements were randomly assigned (60%/40%) to independent derivation (n=84,913) and validation (n=56,584) populations. A sex-specific risk score based on IMRS methods was computed in the derivation population using adjusted multivariable regression weights from all significant and noncollinear CMP [expanded IMRS (eIMRS)] and, when available, WBC differential components (eIMRS+diff).

Results:

Age averaged 67±16 years for females and 67±15 years for males. Receiver operator characteristic (ROC) c-statistics for 30-day death showed marked improvement for the eIMRS compared to the IMRS in both females [0.895 (0.882, 0.908) vs. 0.865 (0.850, 0.880)] and males [0.861 (0.847, 0.876) vs. 0.824 (0.807, 0.841)]. These results persisted for 1-year death: females [0.854 (0.847, 0.862) vs. 0.828 (0.819, 0.836)] and males [0.835 (0.826, 0.844) vs. 0.796 (0.789, 0.808)]. In addition, the eIMRS significantly improved risk reclassification. Further precision was seen when WBC differential components were included.

Conclusions:

The addition of the CMP components to the IMRS improved risk prediction. WBC differential also improved risk score predictive ability. These results suggest that the eIMRS may function even better than IMRS as a tool in patient care, risk-adjustment, and clinical research settings for predicting outcomes.

Association of the dispersion in red blood cell volume with mortality

BACKGROUND

The red cell distribution width (RDW) predicts mortality among many populations. RDW is calculated as the standard deviation (SD) of the red blood cell (RBC) volume divided by mean corpuscular volume (MCV). Because higher MCV also predicts mortality, we hypothesized that the RDW numerator (one SD of RBC volume or 1SD-RDW) predicts mortality more strongly than the RDW.

MATERIAL AND METHODS

Adult subjects hospitalized during a contemporary clinical era (10/2005-1/2014, N = 135,963) and a historical era (1/1999-9/2005, N = 119,530) were studied. The RDW was obtained from the complete blood count (CBC), while 1SD-RDW was calculated (RDW multiplied by MCV and divided by 100).

RESULTS

In univariable Cox regression (2005-2014 cohort), 1SD-RDW (quintile 5 vs. 1: hazard ratio [HR] = 8.38, 95% confidence interval [CI] = 7.94, 8.85; P < 0.001) was a superior predictor of mortality compared to RDW (quintile 5 vs. 1: HR = 4.78, CI = 4.57, 5.00; P < 0.001). This superiority remained after adjustment for age, sex, basic metabolic profile components and other CBC factors excluding MCV (1SD-RDW: HR = 2.41, CI = 2.28, 2.55; RDW: HR = 2.01, CI = 1.92, 2.11). Further adjustment for MCV strengthened the RDW association (HR = 2.14, CI = 2.04, 2.24; P < 0.001), becoming indistinct from 1SD-RDW (HR = 2.20, CI = 2.08, 2.33; P < 0.001). Findings were similar for the 1999-2005 cohort.

CONCLUSIONS

The 1SD-RDW predicted mortality more strongly than RDW, suggesting that 1SD-RDW is superior to RDW as an individual risk predictor. Further, these results indicate that the dispersion of RBC volume and its mean are independent risk markers. Further research is required to understand the clinical value and mechanistic basis of these associations.

Repeated measurement of the intermountain risk score enhances prognostication for mortality

Background

The Intermountain Risk Score (IMRS), composed of the complete blood count (CBC) and basic metabolic profile (BMP), predicts mortality and morbidity in medical and general populations. Whether longitudinal repeated measurement of IMRS is useful for prognostication is an important question for its clinical applicability.

Methods

Females (N = 5,698) and males (N = 5,437) with CBC and BMP panels measured 6 months to 2.0 years apart (mean 1.0 year) had baseline and follow-up IMRS computed. Survival analysis during 4.0±2.5 years (maximum 10 years) evaluated mortality (females: n = 1,255 deaths; males: n = 1,164 deaths) and incident major events (myocardial infarction, heart failure [HF], and stroke).

Results

Both baseline and follow-up IMRS (categorized as high-risk vs. low-risk) were independently associated with mortality (all p<0.001) in bivariable models. For females, follow-up IMRS had hazard ratio (HR) = 5.23 (95% confidence interval [CI] = 4.11, 6.64) and baseline IMRS had HR = 3.66 (CI = 2.94, 4.55). Among males, follow-up IMRS had HR = 4.28 (CI = 3.51, 5.22) and baseline IMRS had HR = 2.32 (CI = 1.91, 2.82). IMRS components such as RDW, measured at both time points, also predicted mortality. Baseline and follow-up IMRS strongly predicted incident HF in both genders.

Conclusions

Repeated measurement of IMRS at baseline and at about one year of follow-up were independently prognostic for mortality and incident HF among initially hospitalized patients. RDW and other CBC and BMP values were also predictive of outcomes. Further research should evaluate the utility of IMRS as a tool for clinical risk adjustment.

The intermountain risk score predicts incremental age-specific long-term survival and life expectancy

The Intermountain Risk Score (IMRS) encapsulates the mortality risk information from all components of the complete blood count (CBC) and basic metabolic profile (BMP), along with age. To individualize the IMRS more clearly, this study evaluated whether IMRS weightings for 1-year mortality predict age-specific survival over more than a decade of follow-up. Sex-specific 1-year IMRS values were calculated for general medical patients with CBC and BMP laboratory tests drawn during 1999-2005. The population was divided randomly 60% (N = 71,921, examination sample) and 40% (N = 47,458, validation sample). Age-specific risk thresholds were established, and both survival and life expectancy were compared across low-, moderate-, and high-risk IMRS categories. During 7.3 ± 1.8 years of follow-up (range, 4.5-11.1 years), the average IMRS of decedents was higher than censored in all age/sex strata (all P < 0.001). For examination and validation samples, every age stratum had incrementally lower survival for higher risk IMRS, with hazard ratios of 2.5-8.5 (P < 0.001). Life expectancies were also significantly shorter for higher risk IMRS (all P < 0.001): For example, among 50-59 year-olds, life expectancy was 7.5, 6.8, and 5.9 years for women with low-, moderate-, and high-risk IMRS (with mortality in 5.7%, 16.3%, and 37.0% of patients, respectively). In Men, life expectancy was 7.3, 6.8, and 5.4 for low-, moderate-, and high-risk IMRS (with patients having 7.3%, 19.5%, and 40.0% mortality), respectively. IMRS significantly stratified survival and life expectancy within age-defined subgroups during more than a decade of follow-up. IMRS may be used to stratify age-specific risk of mortality in research, clinical/preventive, and quality improvement applications. A web calculator is located at http://intermountainhealthcare.org/IMRS.

The Intermountain Risk Score (including the red cell distribution width) predicts heart failure and other morbidity endpoints

AIMS

The complete blood count (CBC) and basic metabolic profile are common, low-cost blood tests, which have previously been used to create and validate the Intermountain Risk Score (IMRS) for mortality prediction. Mortality is the most definitive clinical endpoint, but medical care is more easily applied to modify morbidity and thereby prevent death. This study tested whether IMRS is associated with clinical morbidity endpoints.

METHODS AND RESULTS

Patients seen for coronary angiography (n = 3927) were evaluated using a design similar to a genome-wide association study. The Bonferroni correction for 102 tests required a P-value of ≤ 4.9 × 10⁻⁴ for significance. A second set of angiography patients (n = 10 413) was used to validate significant findings from the first patient sample. In the first patient sample, IMRS predicted heart failure (HF) (P(trend) = 1.6 × 10(-26)), coronary disease (P(trend) = 2.6 × 10(-11)), myocardial infarction (MI) (P(trend) = 3.1 × 10(-25)), atrial fibrillation (P(trend) = 2.5 × 10(-20)), and chronic obstructive pulmonary disease (P(trend) = 4.7 × 10⁻⁴). Even more, IMRS predicted HF readmission [hazard ratio (HR) = 2.29/category, P(trend) = 1.2 × 10⁻⁶), incident HF (HR = 1.88/category, P(trend) = 0.02), and incident MI (HR = 1.56/category, P(trend) = 4.7 × 10⁻⁴). These findings were verified in the second patient sample.

CONCLUSION

Intermountain Risk Score, a predictor of mortality, was associated with morbidity endpoints that often lead to mortality. Further research is required to fully characterize its clinical utility, but its low-cost CBC and basic metabolic profile composition may make it ideal for initial risk estimation and prevention of morbidity and mortality. An IMRS web calculator is freely available at http://intermountainhealthcare.org/IMRS.

Exceptional mortality prediction by risk scores from common laboratory tests

BACKGROUND

Some components of the complete blood count and basic metabolic profile are commonly used risk predictors. Many of their components are not commonly used, but they might contain independent risk information. This study tested the ability of a risk score combining all components to predict all-cause mortality.

METHODS

Patients with baseline complete blood count and basic metabolic profile measurements were randomly assigned (60%/40%) to independent training (N = 71,921) and test (N = 47,458) populations. A third population (N = 16,372) from the Third National Health and Nutrition Examination Survey and a fourth population of patients who underwent coronary angiography (N = 2558) were used as additional validation groups. Risk scores were computed in the training population for 30-day, 1-year, and 5-year mortality using age- and sex-adjusted weights from multivariable modeling of all complete blood count and basic metabolic profile components.

RESULTS

Area under the curve c-statistics were exceptional in the training population for death at 30 days (c = 0.90 for women, 0.87 for men), 1 year (c = 0.87, 0.83), and 5-years (c = 0.90, 0.85) and in the test population for death at 30 days (c = 0.88 for women, 0.85 for men), 1 year (c = 0.86, 0.82), and 5 years (c = 0.89, 0.83). In the test, the Third National Health and Nutrition Examination Survey, and the angiography populations, risk scores were highly associated with death (P <.001), and thresholds of risk significantly stratified all 3 populations.

CONCLUSION

In large patient and general populations, risk scores combining complete blood count and basic metabolic profile components were highly predictive of death. Easily computed in a clinical laboratory at negligible incremental cost, these risk scores aggregate baseline risk information from both the popular and the underused components of ubiquitous laboratory tests.

Superior predictive ability for death of a basic metabolic profile risk score

BACKGROUND

The basic metabolic profile (BMP) is a common blood test containing information about standard blood electrolytes and metabolites. Although individual variables are checked for cardiovascular health and risk, combining them into a total BMP-derived score, as to maximize BMP predictive ability, has not been previously attempted.

METHODS

Patients (N = 279,337) that received a BMP and had long-term follow-up for death were studied. Risk models were created in a training group (60% of study population, n = 167,635), validated in a test group (40% of study population, n = 111,702), and confirmed in the NHANES III (Third National Health and Nutrition Examination Survey) participants (N = 17,752). The BMP models were developed for 30-day, 1-year, and 5-year death using logistic regression with adjustment for age and sex. The BMP parameters were categorized as low, normal, or high based on the standard range of normal. Glucose was categorized as normal, intermediate, and high. Creatinine >or=2 mg/dL was further categorized as very high.

RESULTS

Average age was 53.2 +/- 20.1 years, and 44.3% were male. The areas under the curve for the training and test groups for 30-day, 1-year, and 5-year death were 0.887 and 0.882, 0.850 and 0.848, and 0.858 and 0.847, respectively. The predictive ability of these risk scores was further confirmed in the NHANES III population and independent of the Framingham Risk Score.

CONCLUSION

In large, prospectively followed populations, a highly significant predictive ability for death was found for a BMP risk model. We propose a total BMP score as an optimization of this routine baseline test to provide an important new addition to risk prediction.

Usefulness of a complete blood count-derived risk score to predict incident mortality in patients with suspected cardiovascular disease

The complete blood cell (CBC) count is an inexpensive, frequently obtained blood test whose information content is potentially underused. We examined the predictive ability of the CBC count for incident death in 29,526 consecutive consenting patients who underwent coronary angiography. Subjects were randomly assigned to training (60%) and test (40%) groups and were followed for an average of 4.9 years. Computed and integer risk score models for all-cause death were developed for 30 days and 1, 5, and 10 years using multivariable logistic regressions applied to CBC metrics, age, and gender. The study cohort was an average age of 61 years, 62% were men, and had a 3.3% annual risk of mortality. An integer (scalar) risk score (range 0 to 18) successfully separated patient cohorts into subgroups at markedly different mortality risks (<1% to >14% at 30 days). Predictive fractions (area under risk curve) at 30 days for the CBC-only model and the age- and gender-adjusted CBC model were 0.76 and 0.78, respectively, in the training set and 0.71 and 0.75, respectively, in the test set (all p values <<0.001). The CBC model was markedly more informative than models based only on hematocrit, white blood cell count, or age and gender and was superior to models with all 7 traditional risk factors. In conclusion, in a large, prospectively assembled database, a CBC risk model had high predictive ability for risk of incident mortality. A total CBC score is an important new addition to risk prediction, and it can be easily generated by computer for clinical use at negligible incremental cost.

C-Reactive Protein Predicts Death in Patients with Non-Ischemic Cardiomyopathy

C-reactive protein (CRP) has been associated with atherosclerotic complications, and we hypothesized that CRP levels might also predict death in non-ischemic patients with left ventricular dysfunction. Two hundred and three patients with non-ischemic left ventricular dysfunction undergoing cardiac catheterization were included and were followed for 2.4 +/- 1.4 years to determine the incidence of fatal events. Death occurred in 15% of patients with low CRP (1st and 2nd tertiles) and 30% of patients with high CRP (3rd tertile). After adjustment for 11 covariates, high CRP (p = 0.037, hazard ratio = 2.0) significantly and independently predicted mortality. Even in the absence of coronary artery disease, patients with left ventricular dysfunction are at increased risk of mortality based on their baseline CRP concentrations.