Clinical prediction rule for SARS-CoV-2 infection from 116 U.S. emergency departments 2-22-2021

Objectives

Accurate and reliable criteria to rapidly estimate the probability of infection with the novel coronavirus-2 that causes the severe acute respiratory syndrome (SARS-CoV-2) and associated disease (COVID-19) remain an urgent unmet need, especially in emergency care. The objective was to derive and validate a clinical prediction score for SARS-CoV-2 infection that uses simple criteria widely available at the point of care.

Methods

Data came from the registry data from the national REgistry of suspected COVID-19 in EmeRgency care (RECOVER network) comprising 116 hospitals from 25 states in the US. Clinical variables and 30-day outcomes were abstracted from medical records of 19,850 emergency department (ED) patients tested for SARS-CoV-2. The criterion standard for diagnosis of SARS-CoV-2 required a positive molecular test from a swabbed sample or positive antibody testing within 30 days. The prediction score was derived from a 50% random sample (n = 9,925) using unadjusted analysis of 107 candidate variables as a screening step, followed by stepwise forward logistic regression on 72 variables.

Results

Multivariable regression yielded a 13-variable score, which was simplified to a 13-point score: +1 point each for age>50 years, measured temperature>37.5°C, oxygen saturation<95%, Black race, Hispanic or Latino ethnicity, household contact with known or suspected COVID-19, patient reported history of dry cough, anosmia/dysgeusia, myalgias or fever; and -1 point each for White race, no direct contact with infected person, or smoking. In the validation sample (n = 9,975), the probability from logistic regression score produced an area under the receiver operating characteristic curve of 0.80 (95% CI: 0.79–0.81), and this level of accuracy was retained across patients enrolled from the early spring to summer of 2020. In the simplified score, a score of zero produced a sensitivity of 95.6% (94.8–96.3%), specificity of 20.0% (19.0–21.0%), negative likelihood ratio of 0.22 (0.19–0.26). Increasing points on the simplified score predicted higher probability of infection (e.g., >75% probability with +5 or more points).

Conclusion

Criteria that are available at the point of care can accurately predict the probability of SARS-CoV-2 infection. These criteria could assist with decisions about isolation and testing at high throughput checkpoints.

A new method to incorporate age and gender into the criteria for the detection of acute inferior myocardial infarction

Recent studies have shown that younger women are more likely to die during and after hospitalization for acute myocardial infarction (MI) than older women and men of all ages. This may be partly due to incorrect diagnosis or late detection of acute MI in younger women. At high specificity levels (>98%), the sensitivity of the initial ECG to detect acute MI may be as low as 30% when using traditional criteria by both physicians and computerized interpretation programs. This study examines if women of different age groups have a similar ECG presentation to men during acute inferior MI and if the diagnostic accuracies of the initial ECG are comparable. We analyzed chest pain ECGs from Mayo Clinic and Medical College of Wisconsin, which included 1,339 patients with acute inferior MI and 1,169 age-matched controls with noncardiac chest pain. We subdivided all groups by age (below and above 60 years) and compared ECG parameters (ST elevation, ST depression, QRS duration, R-wave amplitude, Q-wave duration and amplitude, QT interval) between genders. For inferior MI patients under age 60, women had lower ST elevations at the J point in lead II than men (57 +/- 91 microV vs. 86 +/- 117 microV, P < .02). This trend was reversed for patients over age 60 (lead a VF: 102 +/- 126 microV vs. 84+/-117 microV, P < .04; Lead III: 130+/-146 microV vs. 103+/-131 microV, P < .007). A neural network method was used to identify the most significant group of ECG parameters for detecting acute MI. An adaptive fuzzy logic method was developed for adapting to the threshold differences among the different gender and age groups. The new algorithm improved the sensitivity of acute inferior MI detection by more than 25% relative to old algorithm, while maintaining the high specificity around 98% for noncardiac chest pain patients.

Utilization of cardiopulmonary resuscitation in nursing homes in one community: rates and nursing home characteristics

OBJECTIVE

To determine the rate of cardiopulmonary resuscitation use among all nursing homes in a large urban area, to examine CPR use over time, to discover whether CPR use varies among nursing homes, and to describe characteristics of patients undergoing CPR.

DESIGN

Retrospective survey.

SETTING

Nursing homes in a large urban area.

PARTICIPANTS

One hundred ninety-six nursing home residents of 68 nursing homes underwent cardiopulmonary resuscitation over a 4-year period (1986-1989). Over this time there were 9,486 deaths in these homes, which comprised 10,252 beds.

MEASUREMENTS

The CPR:death ratio was determined for each facility. The ratio was analyzed over time and by type of facility (eg, proprietary, non-profit, size of facility). The ratio was also examined among facilities with variable death rates. Patients undergoing CPR are described.

RESULTS

The ratio of CPR:death over the 4-year period was 0.02. CPR:death ratio was higher (0.03) for the proprietary homes compared with the non-profit homes (0.01) P < 0.0001. A significant downward trend of CPR:death was noted over the study period for the non-profit homes; no such trend was noted in the proprietary homes. Size of nursing home did not influence the rate of CPR use. Homes with greater numbers of deaths per bed had a lower utilization of CPR. Patients undergoing CPR were old, frail, and had multiple medical problems. CPR attempts were frequent around the time of nursing home admission.

CONCLUSION

The utilization of CPR in nursing homes is quite low. Non-profit homes utilize CPR less than proprietary homes. Nursing homes with the highest numbers of deaths per bed utilize CPR less than homes with lower numbers of deaths per bed. Nursing home residents receiving CPR are quite old, have multiple illnesses, and are impaired.

Time delays in the diagnosis and treatment of acute myocardial infarction: a tale of eight cities. Report from the Pre-hospital Study Group and the Cincinnati Heart Project

To establish the magnitude of prehospital and hospital delays in initiating thrombolytic therapy for acute myocardial infarction, the time from telephone 911 emergency medical system (EMS) activation to treatment and its components were analyzed from eight separate ongoing trials. This included estimates of ambulance response time, prehospital evaluation and treatment time, and time from admission to the hospital to initiation of thrombolytic therapy. The average time from EMS activation to patient arrival at the hospital was prospectively determined to be 46.1 +/- 8.2 minutes in 3715 patients from eight centers. The time from admission to the hospital to initiation of thrombolytic therapy was retrospectively determined to be 83.8 +/- 55.0 minutes in a separate group of 730 patients from six centers. Both the prehospital and hospital time delays were much longer than those perceived by paramedics and emergency department directors. Shorter hospital time delays were observed in patients in whom a prehospital ECG was obtained as part of a protocol-driven prehospital diagnostic strategy and a diagnosis of acute infarction made before arrival at the hospital (36.3 +/- 11.3 minutes in 13 patients). These results show that the magnitude of time required to evaluate, transport, and initiate thrombolytic therapy will preclude initiation of treatment to most patients within the first hour of symptoms. Implementation of a protocol-driven prehospital diagnostic strategy may be associated with a reduction in time to thrombolytic therapy.

Electromechanical dissociation: six years prehospital experience

Electromechanical dissociation (EMD) is the presenting rhythm in approximately 17% of all prehospital cardiorespiratory arrests. Yet, we know comparatively little about the demographic profile of these patients. The purpose of this study was to review historical and resuscitative parameters to help create a demographic profile. For a 6-year period of time from January 1st, 1980 to December 31st, 1985, 503 adult patients presented to a prehospital system in non-traumatic, nonpoisoned, cardiorespiratory arrest with an initial rhythm of electromechanical dissociation. The overall average response time was 6.1 +/- 3.2 min. Sixty percent of the patients were witnessed arrests and 65% had bystander initiated CPR. Forty-six percent of the patients had a cardiac history: myocardial infarction 13%, CHF 11% and other 21%. Other pertinent past medical history included diabetes 15%, COPD 10% and seizures 3%. The average age was 69.8 +/- 13.7 years. Fifty-seven percent were male. Forty-three percent were on cardiac medication including: digoxin, 24%; nitroglycerin, 12%; potassium supplements, 9%; propranolol, 8%; isordil, 6%; quinidine, 3%; nitropaste, 3%; and other cardiac medications, 15%. One hundred forty-eight (29%) patients developed a pulse at some time during resuscitative efforts, of these 17 (3.4%) patients responded with a pulse immediately after intubation. The mean time of resuscitation to sustaining pulse was 20 +/- 11 min and the mean resuscitation time to sustaining pressure was 22 +/- 11 min. Nineteen percent were successfully resuscitated, defined as a conveyance of a patient with a pulse and a rhythm to an emergency department. Four point four percent were saved, defined as a patient discharged alive from the hospital. Approximately 53% of the successfully resuscitated patients and 45% of the save patients were determined to have a probable respiratory event as the primary etiology of their arrest. This study attempts to provide some insight into the demographic profile of the patients in EMD.

Defining electromechanical dissociation: morphologic presentation

Electromechanical dissociation (EMD) is inconsistently defined in the literature. Our definition is the presence of discernible electrical complexes (excluding ventricular tachycardia and ventricular fibrillation) and the absence of palpable pulses. It has been noted that EMD may present with a variety of morphological complexes. It was the purpose of this study to categorize the electrical morphologic characteristics of patients presenting in EMD and to correlate morphology with patient outcome and response to therapy. From the 6-year period, January 1st, 1980 to December 31st, 1985, 503 evaluable adult patients presented to an urban paramedic system in non-traumatic, non-poisoned, cardiorespiratory arrest and were determined to be in EMD. The rhythm strips obtained from paramedics on all patients were retrospectively reviewed and were arbitrarily categorized in the following manner: Group 1, normal QRS width, isoelectric ST and normal appearing T-waves; Group 2A, atrial activity, widened QRS width (greater than or equal to 0.12 ms) or abnormal ST and/or T-waves (ST depression, elevation, slurring or T-wave inversion); Group 2B, same as Group 2A but without atrial activity; Group 3, essentially monophasic, slurred RST complexes. The respective initial distribution was Group 1, 147 (29%); Group 2A, 248 (49%); Group 2B, 60 (12%); Group 3, 48 (10%). The relative frequency of morphologies preceding the attainment of a pulse was as follows: Group 1, 30 (24%); Group 2A, 82 (65%); Group 2B, 8 (6%); Group 3, 6 (5%) (P less than or equal to 0.01 with no significant difference between Group 2B and 3).(ABSTRACT TRUNCATED AT 250 WORDS)

Interventions in the therapy of electromechanical dissociation

Electromechanical dissociation (EMD) is a major arrest rhythm for which there is often inadequate treatment. The purpose of this study was to evaluate the different pharmacological and non-pharmacological interventions considered in the treatment of EMD. During the 6-year period, January 1st, 1980 to December 31st, 1985, 503 evaluable adult patients presented in a non-traumatic, non-poisoning cardiopulmonary arrest with the initial rhythm of EMD. One hundred nineteen patients obtained a pulse during resuscitation efforts following drug administration. The average time to obtaining pulses after the last drug administration was 1.97 +/- 2.21 min. The following drugs were last administered prior to transient pulses: bicarbonate, 31/119 (26%); epinephrine, 26/119 (22%); atropine, 26/119 (22%); dopamine, 13/119 (11%); calcium, 11/119 (9%); isoproterenol, 7/119 (6%); other drugs, 5/119 (4%). Ninety-five percent of the successful resuscitations received eight or less drug interventions and all saves received three or less drug interventions. Two hundred twenty-four patients (44.5%) had 288 non-pharmacological interventions. Twenty-three patients developed a pulse after intervention in the following distribution: MAST suit (N = 9), pericardiocentesis (N = 6), fluid challenge (N = 5), needle thoracostomy (N = 1), and intervention combinations (N = 2). The time interval between intervention and the onset of pulse was as follows: MAST suit, 4 +/- 2.8 min; pericardiocentesis, 3.7 +/- 3.6 min; fluid challenge, 4.8 +/- 4.1 min; needle thoracostomy, 6 min. The overall save rate for intervention patients was 0.9% whereas for those not having intervention it was 7.2% (P less than or equal to 0.0003).(ABSTRACT TRUNCATED AT 250 WORDS)

Use of carbon monoxide to measure luminal stirring in the rat gut

We used carbon monoxide (CO) as a probe to quantitatively measure intestinal unstirred water layers in vivo. CO has several features that make it uniquely well suited to measure the unstirred layer in that its tight binding to hemoglobin makes uptake diffusion limited, and its relatively high lipid solubility renders membrane resistance negligible relative to the water barriers of the unstirred layer and epithelial cell. The unique application of CO was the measurement of the absorption rate of CO both from the gas phase as well as a solute dissolved in saline. Several lines of evidence showed that a gut stripped free of saline and then filled with gas contained a negligible unstirred layer. Thus, absorption of CO from the gas phase measured resistance of just the epithelial cell. Subtraction of this value from the resistance of CO absorption from saline provided a direct measure of unstirred layer resistance. Studies in the rat showed for a 3-min absorption period that the conventionally calculated apparent unstirred layer for the jejunum was 411 micron and for the colon was 240 micron. However, this conventionally calculated unstirred layer resistance did not truly depict the situation in the rat gut, since there was a continuing depletion of CO from outer surfaces of luminal contents throughout the experiment period. This produced a continually increasing diffusion barrier with time. Calculation of expected absorption rate from unstirred cylinders with the dimensions of the rat gut indicated that there was virtually no stirring in the small intestine and minimal stirring in the colon. The technique described in this paper appears to be simpler and to require fewer assumptions for validity than other techniques previously used to measure unstirred layers in vivo.