Abstract
Objectives There is lack of clarity regarding the role of angiotensin receptor blockers (ARB) or angiotensin converting enzyme inhibitors (ACEi) in interfering with the SARS-COV-2 binding on human cells and the resulting change in disease severity. We sought to assess the risk of hospitalization for COVID-19 and serious complications in current users of ARB or ACEi compared to users of dihydropyridine calcium channel blockers (dhpCCB).
Design Cohort study
Setting The analysis used de-identified, patient level data from HealthVerity, linking longitudinal data from US medical and pharmacy claims, which contain information on inpatient or outpatient diagnoses, procedures and medication dispensing.
Participants We identified patients aged 40+ and free of chronic kidney disease (CKD) who were newly diagnosed COVID-19, between March 1, 2020 and May 30, 2020, and adherent to ACEi, ARB, or dhpCCB therapy.
Interventions Current use of an ACEi, ARB, or dhpCCB.
Main outcome measures We compared the 30-day risk of hospitalization for COVID-19 and serious complications.
Results Of 24,708 patients identified, 7,571 were current users of an ARB, 8,484 of an ACEi, and 8,653 of a dhpCCB. The unadjusted 30-day risk of hospitalization for COVID-19 was 2.66% among ARB users, and 2.90% among ACEi users and 3.68% in dhpCCB users. In the PS-matched cohort, the risk of hospitalization among ARB users was 17% lower as compared to dhpCCB (RR=0.83; 0.68-1.00), and the risk among ACE users was 10% lower as compared to dhpCCB (RR=0.90; 0.76-1.07). When including patients with pre-existing CKD, the protective effect of ARB (RR= 0.74; 0.62-0.88) and ACEi (RR=0.84; 0.71-0.99) was more pronounced.
Conclusions This cohort study showed that neither ARB nor ACEi use increase the risk of severe COVID-19 disease among those infected, and instead suggests that current use of ARB may offer a protective effect. This study found no evidence to support the discontinuation of ARB/ACEi therapy.
Background
The spike proteins covering SARS-COV-2 bind to angiotensin-converting enzyme-2 (ACE2) receptors primarily on type II alveolar cells and cells in the heart, kidney, liver, and gastrointestinal tract, enabling the virus to inject its RNA. 1 It has been hypothesized that blocking the ACE2 receptor may decrease SARS-COV-2’s ability to bind to human cells, and thus reduce the severity of COVID-19 illness.1 However, there is uncertainty surrounding the effect of renin-angiotensin-aldosterone system (RAAS) inhibition in patients with COVID-19, with some hypothesizing that angiotensin receptor blockers (ARB) or angiotensin converting enzyme inhibitors (ACEi) use may increase the propagation of the virus. 1-4
Any link between ACEi or ARB use and the virus RNA’s ability to enter human cells could yield an effect, positive or negative, on the clinical severity of a COVID-19 infection, including the need for hospitalization, ventilator use, intensive care unit (ICU) stay, or death. Pre-published cohort studies by Giorgi, Mehta, and Rentsch examined the use of ACEi/ARB with respect to hospital admissions and found an increased risk; however, the contrast they made compared ACEi/ARB users to non-users and likely did not fully account for underlying factors associated with both the need for ACE or ARB treatment and with poor health outcomes.5-7 Khera et al. addressed some of these issues by equalizing risk through requiring comparator patients to have hypertension and to be treated with a guideline-recommended antihypertensive, and found a protective effect with ACEi use, but not ARB use, among Medicare beneficiaries.8 The literature lacks study findings examining the use of ACEi or ARB therapy across a broad population, and the associated risk of severe outcomes from COVID-19.4 As such, clinicians lack clear guidance on appropriate action to take with patients using an ACEi or ARB and diagnosed with COVID-19, in particular whether to discontinue therapy.
To that end, this study compares the risk of hospitalization for COVID-19 disease and hospitalization for serious complications of COVID-19 among prevalent users of ACEi or ARB as compared with prevalent users of dihydropyridine calcium channel blockers (dhpCCB), using recently-collected administrative claims data from the US.
Methods
Data source
This study used de-identified patient level healthcare data from HealthVerity covering >100 million unique patients with data from over 60 unique data sources in the US, including medical claims, pharmacy claims, lab, and electronic medical records (EMR). Data encompass all major US payer types, including commercial, Medicaid, and Medicare. For this study, we examined linked medical claims, pharmacy claims, and laboratory data containing diagnoses, procedures, prescriptions and other variables. Claims included both unadjudicated insurance claims available in near real-time time from practice management systems, billing systems, and claims clearinghouses which were then linked with adjudicated claims sourced from insurance providers and payers as well as laboratory data from hospital billing systems. Data for this study covered the period of December 1, 2018 through May 30, 2020. Claims include dates of service and diagnoses, and are recorded longitudinally, thus allowing us to establish baseline health status, COVID-19 diagnosis, and any sequalae. This study was approved by the New England IRB (#1-9757-1). De-identified patient level study data can be made available for inspection.
Patients
Eligible patients had at least one medical encounter related to COVID-19. A COVID-19-related encounter was defined as any claim with an International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) code in accordance with the Center for Disease Control’s (CDC) interim (B34.2, B97.21, B97.29, J12.81) and final (U07.1) coding guidance for confirmed cases of COVID-19a or who had a positive or presumed positive COVID-19 viral lab test between March 1, 2020 and May 30, 2020 (see Appendix A for further details).
To compare the impact of ongoing ACEi or ARB therapy on severity of COVID-19 infection, we implemented a prevalent user, active comparator design where balance between treatment groups at baseline was achieved through study design and propensity score (PS) matching.9,10 A prevalent user design, where all study participants were using one of the study medications prior to the index date, was chosen in order to understand the effect of medications being taken at the time of COVID-19 infection, rather than the effect of initiation of therapy in response to COVID-19 diagnosis. To improve comparability among treatment groups, prevalent users of a dhpCCB were selected as the active comparator; these agents are also an anti-hypertensive treatment, like ACEi and ARB, yet due to their different mechanism of action, have no hypothesized effect on COVID-19.
We created 2 cohorts for analysis: cohort 1 included prevalent users of ARB versus dhpCCB, and cohort 2 included prevalent users of ACEi versus dhpCCB. Cohort entry was defined as the first outpatient visit with a COVID-19 diagnosis among patients with at least one prescription for ana CDC. ACEi, ARB, or dhpCCB within the 90 days before cohort entry. We excluded patients if they: (1) had no other medical or pharmacy claims activity in the 90 days prior to cohort entry, (2) qualified for more than one treatment category, (3) were of age less than 40 years on cohort entry, or (4) had a diagnosis of chronic kidney disease (CKD) within the 90 days prior to cohort entry, as CKD is a contraindication for use of an ACEi or ARB (though not for dhpCCB). We conducted a sensitivity analysis that included patients with pre-existing CKD.
Outcomes
We compared the risk of (1) hospitalization for COVID-19 and (2) hospitalization for serious COVID-19 complications in the 30 days following cohort entry. We identified hospitalization for COVID-19 as the first inpatient visit with a diagnosis of COVID-19 (ICD-10 code: U07.1) following CDC-defined outpatient diagnosis of COVID-19. We identified hospitalization for serious COVID-19 complications as the first inpatient visit with a diagnosis of COVID-19, as above, followed on the same day or within 30 days by a diagnosis code for acute organ dysfunction or failure, acute respiratory distress syndrome (ARDS), sepsis, or respiratory intubation or mechanical ventilation (see Appendix B for a full list of codes).
Follow-up started the day after cohort entry and continued until the earliest of occurrence of outcome, death, disenrollment, 30 days, or the end of data (May 30, 2020).
Patient characteristics
For each patient, we recorded age at cohort entry, sex, and month of cohort entry. Over the 90 days prior to cohort entry, we assessed: history of pneumonia infection requiring an office visit or hospitalization, prior use of asthma medications or anticoagulants, number of physician visits, number of unique prescription medications filled, and a range of comorbidities that may be risk factors for more severe COVID-19 disease (Table 2). Comorbidities included malignancy, alcohol use disorder, coagulation deficiency, HIV or AIDS, paralysis, valvular heart disease, weight loss, overweight or obesity, smoking, acute coronary syndrome, arrythmias, asthma, chronic pulmonary disease, conduction disorders, congestive heart failure, dependence on supplemental oxygen, hematologic disorders, hypertension, inflammatory heart disease, ischemic heart disease, liver disease, myocardial infarction, stroke, pulmonary circulation disorders, rheumatic disease, and systemic lupus erythematosus.
Statistical analysis
We tabulated baseline patient characteristics and assessed the 30-day risk of the outcomes of interest with 95% confidence intervals for each exposure group.
To account for potential confounding by measured patient characteristics, we used propensity score (PS)-matching to achieve balance of covariates between the treatment groups.11,12 While a prevalent user analysis can lead to bias from adjusting for intermediate variables,13 in this case both the drug exposure and the patient characteristics were observed before cohort entry, and thus are unlikely to cause bias. The PS, defined as the probability that a patient was using either ACEi/ARB versus dhpCCB, conditioning on all the confounders listed above, was estimated with a multivariable logistic regression model. PS-matching was performed using 1:1 nearest neighbor matching with a maximum matching caliper of 0.02 on the PS scale.14 To quantify balance achieved at baseline, we computed standardized differences in the covariates between the two groups after PS-matching.15
We calculated relative risks (RR) and 95% confidence intervals (CIs) of the outcomes associated with drug use using logistic regression. All analyses were conducted using the Aetion Evidence Platform v3.19 (including R version 4.0.2), which has been validated and has successfully predicted randomized control trial findings.16-18 All analyses from the raw patient-level data are recorded with audit trails and can be reviewed on request.
Results
We identified 24,708 patients aged 40 years or older who had an outpatient diagnosis of COVID-19 and met the inclusion criteria (Table 1). At the start of follow-up, cohort 1 had 15,829 prevalent users of either an ARB (n=7,571) or dhpCCB (n=8,258), while cohort 2 had 17,137 prevalent users of either an ACEi (n=8,484) or dhpCCB (n=8,653) (Tables 2, 3). After applying 1:1 PS-matching on all baseline characteristics, cohort 1 had 6,710 pairs of patients using ARB vs. dhpCCB and cohort 2 had 7,334 pairs of patients using ACEi vs. dhpCCB. In the PS-matched cohort, all 60 baseline characteristics were well-balanced between treatment groups (Tables 2, 3) with all absolute standardized differences between treatment groups below the recognized threshold of 0.1.19
In the PS-matched cohort, the 30-day risk of hospitalization for COVID-19 was 2.85% in ARB users (vs. 3.44% in dhpCCB users) and 3.12% in ACEi users (vs. 3.46% in dhpCCB users) (Table 4). After PS-matching we found that ARB users had a 17% decrease in the risk of hospitalization for COVID-19 (RR=0.83; 95% CI, 0.68-1.00) as compared to dhpCCB, while ACEi users had a 10% decrease in risk for hospitalization for COVID-19 (RR=0.90; 0.76-1.07) as compared to dhpCCB. Point estimates suggested protective effects for several serious complications, though confidence intervals generally included the null: acute organ dysfunction or failure (RR=0.88 [95% CI: 0.65-1.20] for ARB and 1.00 [0.76-1.32] for ACEi); ARDS (RR=0.86 [0.63-1.16] for ARB and 0.99 [0.75-1.31] for ACEi); sepsis (RR=0.71 [0.49-1.03] for ARB and 1.06 [0.76-1.48] for ACEi); and for respiratory intubation or mechanical ventilation (RR=0.77 [0.59-1.00] for ARB and 0.99 [0.78-1.26] for ACEi) (Table 4).
In our sensitivity analysis including patients with pre-existing CKD in the 90 days before outpatient COVID-19 diagnosis, the relative risk of hospitalization for COVID-19 decreased to 0.74 (0.62-0.88) for ARB and 0.84 (0.71-0.99) for ACEi.
Discussion
In this population-representative study of patients aged 40 and older with an outpatient COVID-19 diagnosis, we observed a 17% decrease in the 30-day risk of hospitalization for COVID-19 among ARB users compared to dhpCCB users, and a 10% decrease in risk of hospitalization for COVID-19 among ACEi users compared to dhpCCB users. We also observed a numerically decreased 30-day risk of respiratory intubation and mechanical ventilation after hospitalization with COVID-19 among ARB users compared to dhpCCB users. There was no meaningful difference in the risk of acute organ dysfunction, ARDS, or sepsis among RAAS inhibitors vs. dhpCCBs.
Our findings suggest neither a need to discontinue ACEi/ARB therapy among patients using those medications at the time of COVID-19 diagnosis, nor a clear case for use of ACEi/ARB therapy as prophylaxis among COVID-19-infected patients.
There were several notable strengths of this study. First, we used a large population-representative cohort with over 20,000 COVID-19 adults 40+ receiving anti-hypertensive treatment with RAAS inhibitors or dhpCCB from established data sources. Second, the outcome of interest was severe enough to avoid differential surveillance bias since all serious infections that require hospitalization are captured in claims data, and the ICD codes used to identify them have been shown to be of high accuracy.20 Third, we implemented robust adjustment for potential confounders using active comparator study design combined with 1:1 propensity score matching.21,22 Fourth, by requiring comparator patients also be treated with an antihypertensive medication, the confounder-adjusted, dhpCCB control group provided an appropriate reference value for hospitalization in patients aged 40 or older with COVID-19 infection.
This population-based study applying a causal study design concludes that neither ARB nor ACEi use increase the risk of severe COVID-disease among those infected, and rather suggest that prevalent use of these medications – ARB in particular -- may offer a protective effect. Indeed, our findings suggest a small decrease in risk of hospitalization for COVID-19 between adult patients with an outpatient COVID-19 diagnosis who were using RAAS inhibitors versus those treated with dhpCCB. These findings are reassuring as they suggest that RAAS inhibitors should not be discontinued among patients with COVID-19 infection.
Data Availability
De-identified patient level study data can be made available for inspection. All analyses from the raw patient-level data are recorded with audit trails and can be reviewed on request.
Appendix materials
Abbreviations: ARB, angiotensin receptor blocker; ACEi, angiotensin converting enzyme inhibitor; dhpCCB, dihydropyridine calcium channel blocker; CKD, chronic kidney disease.
*Follow-up started the day after cohort entry until the earliest occurrence of a censoring event: outcome, 30 days after cohort entry, death, disenrollment, or end of study period (May 30, 2020).
Appendix A Coding algorithms for population and exposure
REFERENT: dhpCCB (Excludes: verapamil and diltiazem).
The occurrence of Pharmacy Claims with the following attributes:
Generic Name is any of: { “CLEVIDIPINE BUTYRATE”, “ALISKIREN HEMIFUMARATE/AMLODIPINE/HYDROCHLOROTHIAZIDE”, “MIBEFRADIL DI-HCL”, “ALISKIREN HEMIFUMARATE/AMLODIPINE BESYLATE”, “BEPRIDIL HCL”, “NICARDIPINE IN DEXTROSE, ISO-OSMOTIC”, “NICARDIPINE IN SODIUM CHLORIDE, ISO-OSMOTIC”, “NIMODIPINE”, “ISRADIPINE”, “NICARDIPINE HCL”, “AMLODIPINE BESYLATE/ATORVASTATIN CALCIUM”, “FELODIPINE”, “AMLODIPINE BESYLATE”, “NIFEDIPINE” }
CLEVIDIPINE BUTYRATE
MIBEFRADIL DI-HCL
BEPRIDIL HCL
NICARDIPINE IN DEXTROSE, ISO-OSMOTIC
NICARDIPINE IN SODIUM CHLORIDE, ISO-OSMOTIC
NIMODIPINE
ISRADIPINE
NICARDIPINE HCL
AMLODIPINE BESYLATE/ATORVASTATIN CALCIUM
FELODIPINE
AMLODIPINE BESYLATE
NIFEDIPINE
EXPOSURE: ACEi
The occurrence of Pharmacy Claims with the following attributes:
Generic Name is any of: { “AMLODIPINE BESYLATE/BENAZEPRIL HCL”, “BENAZEPRIL HCL”, “BENAZEPRIL HCL/HYDROCHLOROTHIAZIDE”, “CAPTOPRIL”, “CAPTOPRIL/HYDROCHLOROTHIAZIDE”, “ENALAPRIL MALEATE”, “ENALAPRIL MALEATE/DILTIAZEM MALATE”, “ENALAPRIL MALEATE/FELODIPINE”, “ENALAPRIL MALEATE/HYDROCHLOROTHIAZIDE”, “ENALAPRILAT DIHYDRATE”, “FOSINOPRIL SODIUM”, “FOSINOPRIL SODIUM/HYDROCHLOROTHIAZIDE”, “LISINOPRIL”, “LISINOPRIL/DIETARY SUPPLEMENT,COMB.10”, “LISINOPRIL/HYDROCHLOROTHIAZIDE”, “MOEXIPRIL HCL”, “MOEXIPRIL HCL/HYDROCHLOROTHIAZIDE”, “PERINDOPRIL ARGININE/AMLODIPINE BESYLATE”, “PERINDOPRIL ERBUMINE”, “QUINAPRIL HCL”, “QUINAPRIL HCL/HYDROCHLOROTHIAZIDE”, “RAMIPRIL”, “TRANDOLAPRIL”, “TRANDOLAPRIL/VERAPAMIL HCL” }
BENAZEPRIL HCL
BENAZEPRIL HCL/HYDROCHLOROTHIAZIDE
CAPTOPRIL
CAPTOPRIL/HYDROCHLOROTHIAZIDE
ENALAPRIL MALEATE
ENALAPRIL MALEATE/HYDROCHLOROTHIAZIDE
ENALAPRILAT DIHYDRATE
FOSINOPRIL SODIUM
FOSINOPRIL SODIUM/HYDROCHLOROTHIAZIDE
LISINOPRIL
LISINOPRIL/DIETARY SUPPLEMENT,COMB.10
LISINOPRIL/HYDROCHLOROTHIAZIDE
MOEXIPRIL HCL
MOEXIPRIL HCL/HYDROCHLOROTHIAZIDE
PERINDOPRIL ERBUMINE
QUINAPRIL HCL
QUINAPRIL HCL/HYDROCHLOROTHIAZIDE
RAMIPRIL
TRANDOLAPRIL
The occurrence of Pharmacy Claims with the following attributes:
Brand Name is any of: { “ACCUPRIL”, “ACEON”, “ALTACE”, “CAPOTEN”, “LOTENSIN”, “LOTENSIN HCT”, “MAVIK”, “MONOPRIL”, “MONOPRIL HCT”, “PRINIVIL”, “UNIVASC”, “VASOTEC”, “VASOTEC I.V.“, “ZESTRIL” }
ACCUPRIL
ACEON
ALTACE
CAPOTEN
LOTENSIN
LOTENSIN HCT
MAVIK
MONOPRIL
MONOPRIL HCT
PRINIVIL
UNIVASC
VASOTEC
VASOTEC I.V.
ZESTRIL
EXPOSURE: ARB
The occurrence of Pharmacy Claims with the following attributes:
Generic Name is any of: { “ALISKIREN/VALSARTAN”, “AMLODIPINE BESYLATE/OLMESARTAN MEDOXOMIL”, “AMLODIPINE BESYLATE/VALSARTAN”, “AMLODIPINE BESYLATE/VALSARTAN/HYDROCHLOROTHIAZIDE”, “AZILSARTAN MEDOXOMIL”, “AZILSARTAN MEDOXOMIL/CHLORTHALIDONE”, “CANDESARTAN CILEXETIL”, “CANDESARTAN CILEXETIL/HYDROCHLOROTHIAZIDE”, “EPROSARTAN MESYLATE”, “EPROSARTAN MESYLATE/HYDROCHLOROTHIAZIDE”, “IRBESARTAN”, “IRBESARTAN/HYDROCHLOROTHIAZIDE”, “LOSARTAN POTASSIUM”, “LOSARTAN POTASSIUM/HYDROCHLOROTHIAZIDE”, “NEBIVOLOL HCL/VALSARTAN”, “OLMESARTAN MEDOXOMIL”, “OLMESARTAN MEDOXOMIL/AMLODIPINE BESYLATE/HYDROCHLOROTHIAZIDE”, “OLMESARTAN MEDOXOMIL/HYDROCHLOROTHIAZIDE”, “SACUBITRIL/VALSARTAN”, “TELMISARTAN”, “TELMISARTAN/AMLODIPINE BESYLATE”, “TELMISARTAN/HYDROCHLOROTHIAZIDE”, “VALSARTAN”, “VALSARTAN/HYDROCHLOROTHIAZIDE” }
ALISKIREN/VALSARTAN
AZILSARTAN MEDOXOMIL
AZILSARTAN MEDOXOMIL/CHLORTHALIDONE
CANDESARTAN CILEXETIL
CANDESARTAN CILEXETIL/HYDROCHLOROTHIAZIDE
EPROSARTAN MESYLATE
EPROSARTAN MESYLATE/HYDROCHLOROTHIAZIDE
IRBESARTAN
IRBESARTAN/HYDROCHLOROTHIAZIDE
LOSARTAN POTASSIUM
LOSARTAN POTASSIUM/HYDROCHLOROTHIAZIDE
NEBIVOLOL HCL/VALSARTAN
OLMESARTAN MEDOXOMIL
OLMESARTAN MEDOXOMIL/HYDROCHLOROTHIAZIDE
SACUBITRIL/VALSARTAN
TELMISARTAN
TELMISARTAN/HYDROCHLOROTHIAZIDE
VALSARTAN
VALSARTAN/HYDROCHLOROTHIAZIDE
The occurrence of Pharmacy Claims with the following attributes:
Brand Name is any of: { “ATACAND”, “ATACAND HCT”, “AVALIDE”, “AVAPRO”, “BENICAR”, “BENICAR HCT”, “COZAAR”, “DIOVAN”, “DIOVAN HCT”, “EDARBI”, “EXFORGE”, “EXFORGE HCT”, “HYZAAR”, “MICARDIS”, “MICARDIS HCT”, “TEVETEN”, “TEVETEN HCT” }
ATACAND
ATACAND HCT
AVALIDE
AVAPRO
BENICAR
BENICAR HCT
COZAAR
DIOVAN
DIOVAN HCT
EDARBI
EXFORGE
EXFORGE HCT
HYZAAR
MICARDIS
MICARDIS HCT
TEVETEN
TEVETEN HCT
COVID-19 diagnosis:
This measure creates a logical OR of events, and identifies all occurrences of one or more specified events:
COVID-19, virus identified, outpatient (U07.1)
SARS-CoV-2 viral lab test (labs) with positive or presumptive positive result
Other coronavirus infection (not U07.1)
COVID-19, virus identified:
The occurrence of Medical Claims with the following attributes:
Diagnosis Code, ICD-10 is any of: { “U07.1” }
U07.1 - COVID-19, virus identified
COVID-19 viral lab test, with positive of presumed positive result:
The occurrence of Lab Tests with the following attributes:
Test Ordered Name is any of: { “COVID-19 Nasopharynx”, “COVID-19 Oropharynx”, “COVID-19 Pooled NP/OP”, “SARS CoV 2 RNA(COVID 19), QL NAAT”, “SARS Coronavirus with CoV-2 RNA, Quant RT-PCR”, “SARS-CoV-2 RNA,QL REAL-TIME RT-PCR (COVID-19)“, “SARS-CoV-2 RNA,QL REAL-TIME RT-PCR (COVID-19) - Non-swab”, “COVID-19 Nasal/Nasopharynx”, “SARS CoV 2 RNA, QL Real Time RT PCR”, “SARS CoV 2 RNA, RT-PCR”, “SARS-COV-2 RNA (COVID-19), QUALITATIVE NAAT”, “SARS-CoV-2 RNA, QL, RT PCR (COVID-19) - Swabs”, “SARS-CoV-2 RNA, QL, RT PCR (COVID-19) Swabs”, “COVID-19 Pooled N/NP/OP”, “SARS CORONAVIRUS WITH COV-2 RNA, QUALITATIVE REAL-TIME RTPCR”, “SARS COV 2 RNA, RT PCR”, “SARS CoV 2 RNA(COVID 19), QUALITATIVE NAAT”, “SARS COV W/COV 2 RNA,PCR”, “SARS CoV 2 RNA, QL REAL TIME RT PCR”, “SARS Coronavirus w/CoV 2 RNA,QL Real Time RT PCR”, “SARS-CoV-2 RNA,QL Real-Time RT-PCR (CoVID-19) - Non-swab”, “SARS CORONAVIRUS W/CoV 2 RNA,QL REAL TIME RT PCR”, “SARS-COV-2 RNA, QL REAL-TIME RT-PCR”, “SARS-CoV-2 RNA (COVID-19), Qualitative NAAT”, “SARS COV 2 RNA, QL NAAT”, “SARS Cov 2 RNA, QL NAAT” }
COVID-19 Nasopharynx
COVID-19 Oropharynx
COVID-19 Pooled NP/OP
SARS CoV 2 RNA(COVID 19), QL NAAT
SARS Coronavirus with CoV-2 RNA, Quant RT-PCR
SARS-CoV-2 RNA,QL REAL-TIME RT-PCR (COVID-19)
SARS-CoV-2 RNA,QL REAL-TIME RT-PCR (COVID-19) - Non-swab
COVID-19 Nasal/Nasopharynx
SARS CoV 2 RNA, QL Real Time RT PCR
SARS CoV 2 RNA, RT-PCR
SARS-COV-2 RNA (COVID-19), QUALITATIVE NAAT
SARS-CoV-2 RNA, QL, RT PCR (COVID-19) - Swabs
SARS-CoV-2 RNA, QL, RT PCR (COVID-19) Swabs
COVID-19 Pooled N/NP/OP
SARS CORONAVIRUS WITH COV-2 RNA, QUALITATIVE REAL-TIME RTPCR
SARS COV 2 RNA, RT PCR
SARS CoV 2 RNA(COVID 19), QUALITATIVE NAAT
SARS COV W/COV 2 RNA,PCR
SARS CoV 2 RNA, QL REAL TIME RT PCR
SARS Coronavirus w/CoV 2 RNA,QL Real Time RT PCR
SARS-CoV-2 RNA,QL Real-Time RT-PCR (CoVID-19) - Non-swab
SARS CORONAVIRUS W/CoV 2 RNA,QL REAL TIME RT PCR
SARS-COV-2 RNA, QL REAL-TIME RT-PCR
SARS-CoV-2 RNA (COVID-19), Qualitative NAAT
SARS COV 2 RNA, QL NAAT
SARS Cov 2 RNA, QL NAAT
Result is any of: { “Positive for 2019-nCoV”, “Presumptive Pos. for 2019-nCoV”, “Presumptive Positive 2019-nCoV”, “Presumptive Positive for 2019-nCoV” }
Positive for 2019-nCoV
Presumptive Pos. for 2019-nCoV
Presumptive Positive 2019-nCoV
Presumptive Positive for 2019-nCoV
The occurrence of Lab Tests with the following attributes:
Result Name is any of: { “COVID-19 SWB”, “SARS CoV 2 RNA, QL Real Time RT PCR”, “SARS Coronavirus w/CoV 2 RNA,QL Real Time RT PCR”, “COVID-19 Nasopharynx”, “SARS-CoV-2 RNA”, “SARS-CoV-2 RNA,QL REAL-TIME RT-PCR (COVID-19)“, “SARS-CoV-2 RNA,QL REAL-TIME RT-PCR (COVID-19) - Non-swab”, “COVID-19 Nasal/Nasopharynx”, “COVID-19 Oropharynx”, “COVID-19 Pooled N/NP/OP”, “COVID-19 Pooled NP/OP”, “SARS CoV 2 RNA(COVID 19), QL NAAT”, “SARS CoV 2 RNA:“, “SARS CoV 2 RNA”, “SARS-CoV-2 RNA,QL Real-Time RT-PCR (CoVID-19) - Non-swab”, “SARS COV 2 RNA, RT PCR”, “SARS CoV 2 RNA, RT PCR”, “SARS-COV-2 RNA:“, “SARS-CoV-2 RNA:” }
COVID-19 SWB
SARS CoV 2 RNA, QL Real Time RT PCR
SARS Coronavirus w/CoV 2 RNA,QL Real Time RT PCR
COVID-19 Nasopharynx
SARS-CoV-2 RNA
SARS-CoV-2 RNA,QL REAL-TIME RT-PCR (COVID-19)
SARS-CoV-2 RNA,QL REAL-TIME RT-PCR (COVID-19) - Non-swab
COVID-19 Nasal/Nasopharynx
COVID-19 Oropharynx
COVID-19 Pooled N/NP/OP
COVID-19 Pooled NP/OP
SARS CoV 2 RNA(COVID 19), QL NAAT
SARS CoV 2 RNA:
SARS CoV 2 RNA
SARS-CoV-2 RNA,QL Real-Time RT-PCR (CoVID-19) - Non-swab
SARS COV 2 RNA, RT PCR
SARS CoV 2 RNA, RT PCR
SARS-COV-2 RNA:
SARS-CoV-2 RNA:
Result is any of: { “Presumptive Positive for 2019-nCoV”, “Presumptive Pos. for 2019-nCoV”, “Positive for 2019-nCoV”, “Presumptive Positive 2019-nCoV” }
Presumptive Positive for 2019-nCoV
Presumptive Pos. for 2019-nCoV
Positive for 2019-nCoV
Presumptive Positive 2019-nCoV
The occurrence of Lab Tests with the following attributes:
LOINC Code is any of: { “943092”, “945006” }
943092
945006
Result is any of: { “Presumptive Positive for 2019-nCoV”, “Presumptive Pos. for 2019-nCoV”, “Positive for 2019-nCoV”, “Presumptive Positive 2019-nCoV” }
Presumptive Positive for 2019-nCoV
Presumptive Pos. for 2019-nCoV
Positive for 2019-nCoV
Presumptive Positive 2019-nCoV
The occurrence of Lab Tests with the following attributes:
Procedure Code, HCPCS and CPT is any of: { “U0002”, “87635”, “U0001” }
U0002 - Non-CDC laboratory test for SARS-CoV-2/2019-nCoV (COVID-19)
87635 - Infectious agent detection by nucleic acid (DNA or RNA); severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (Coronavirus disease [COVID-19]), amplified probe technique
U0001 - CDC laboratory test for SARS-CoV-2/2019-nCoV (COVID-19)
Result is any of: { “Presumptive Positive for 2019-nCoV”, “Presumptive Pos. for 2019-nCoV”, “Positive for 2019-nCoV”, “Presumptive Positive 2019-nCoV” }
Presumptive Positive for 2019-nCoV
Presumptive Pos. for 2019-nCoV
Positive for 2019-nCoV
Presumptive Positive 2019-nCoV
Other coronavirus infection (not U07.1):
This measure creates a logical OR of events, and identifies all occurrences of one or more specified events:
B34.2 - Coronavirus infection, unspecified
B97.29 - Other coronavirus as the cause of diseases classified elsewhere
B97.21 - SARS-associated coronavirus as the cause of diseases classified elsewhere
J12.81 - Pneumonia due to SARS-associated coronavirus
Appendix B Coding algorithms for outcomes
Hospitalization with COVID-19 diagnosis:
The occurrence of Medical Claims with the following attributes:
Diagnosis Code, ICD-10 is any of: { “U07.1” }
U07.1 - COVID-19, virus identified
Inpatient Indicator is any of:
Yes
The occurrence of Medical Claims with the following attributes:
Diagnosis Code, ICD-10 is any of: { “U07.1” }
U07.1 - COVID-19, virus identified
Hospital Service Revenue Code is any of: { “0012”, “0100”, “0110”, “0116”, “0123”, “0125”, “0129”, “0132”, “0144”, “0145”, “0152”, “0154”, “0159”, “0118”, “0121”, “0124”, “0127”, “0130”, “0134”, “0136”, “0139”, “0140”, “0150”, “0158”, “0160”, “0167”, “0987”, “0115”, “0117”, “0119”, “0133”, “0138”, “0141”, “0143”, “0147”, “0148”, “0153”, “0155”, “0164”, “0112”, “0113”, “0114”, “0120”, “0122”, “0126”, “0128”, “0131”, “0135”, “0151”, “0157”, “0169”, “0101”, “0111”, “0142”, “0156” }}
Hospitalization with COVID-19 + the occurrence of ARDS:
This measure was defined as:
1.1. The occurrence of Hospitalization with COVID
1.2. The occurrence of ARDS, acute respiratory failure, and arrest, starting at least 0 days after the start of 1.1, and starting at most 30 days after the start of 1.1.
The event of interest in this group is 1.2.
The occurrence of ARDS, acute respiratory failure, and arrest
The occurrence of Medical Claims with the following attributes:
Diagnosis Code, ICD-10 is any of: { “R09.2”, “J80”, “J96.02”, “J96.20”, “J96.21”, “J96.22”, “J96.00”, “J96.2”, “J96.0”, “J96.01” }
R09.2 - Respiratory arrest
J80 - Acute respiratory distress syndrome
J96.02 - Acute respiratory failure with hypercapnia
J96.20 - Acute and chronic respiratory failure, unspecified whether with hypoxia or hypercapnia
J96.21 - Acute and chronic respiratory failure with hypoxia
J96.22 - Acute and chronic respiratory failure with hypercapnia
J96.00 - Acute respiratory failure, unspecified whether with hypoxia or hypercapnia
J96.2 - Acute and chronic respiratory failure
J96.0 - Acute respiratory failure
J96.01 - Acute respiratory failure with hypoxia
Hospitalization for COVID-19 + the occurrence of Acute organ dysfunction or failure:
This measure was defined as:
1.1. The occurrence of Hospitalization with COVID
1.2. The occurrence of Inpatient hospitalization with acute organ dysfunction or failure, starting at least 0 days after the start of 1.1, and starting at most 30 days after the start of 1.1.
The event of interest in this group is 1.2.
The occurrence of Acute organ dysfunction or failure:
The occurrence of Medical Claims with the following attributes:
Diagnosis Code, ICD-10 is any of: { “G62.81”, “N17.9”, “G93.41”, “J96.0”, “J96.01”, “K72.01”, “N17”, “N17.1”, “D65”, “K72.0”, “K72.00”, “N17.2”, “J96.00”, “J96.02”, “N17.8”, “G72.81”, “N17.0” }}
G62.81 - Critical illness polyneuropathy
N17.9 - Acute kidney failure, unspecified
G93.41 - Metabolic encephalopathy
J96.0 - Acute respiratory failure
J96.01 - Acute respiratory failure with hypoxia
K72.01 - Acute and subacute hepatic failure with coma
N17 - Acute kidney failure
N17.1 - Acute kidney failure with acute cortical necrosis
D65 - Disseminated intravascular coagulation [defibrination syndrome]
K72.0 - Acute and subacute hepatic failure
K72.00 - Acute and subacute hepatic failure without coma
N17.2 - Acute kidney failure with medullary necrosis
J96.00 - Acute respiratory failure, unspecified whether with hypoxia or hypercapnia
J96.02 - Acute respiratory failure with hypercapnia
N17.8 - Other acute kidney failure
G72.81 - Critical illness myopathy
N17.0 - Acute kidney failure with tubular necrosis
Inpatient Indicator is any of:
Yes
The occurrence of Medical Claims with the following attributes:
Diagnosis Code, ICD-10 is any of: { “G62.81”, “N17.9”, “G93.41”, “J96.0”, “J96.01”, “K72.01”, “N17”, “N17.1”, “D65”, “K72.0”, “K72.00”, “N17.2”, “J96.00”, “J96.02”, “N17.8”, “G72.81”, “N17.0” }
Suspected Inpatient Indicator is any of:
Yes
Hospitalization with COVID-19 + the occurrence of Sepsis:
This measure was defined as:
1.1. The occurrence of Hospitalization with COVID
1.2. The occurrence of Occurrence of sepsis inpatient, starting at least 0 days after the start of 1.1, and starting at most 30 days after the start of 1.1.
The event of interest in this group is 1.2.
The occurrence of sepsis, inpatient:
The occurrence of Medical Claims with the following attributes:
Diagnosis Code, ICD-10 is any of: { “A02.1”, “A26.7”, “A40”, “A40.1”, “A41.01”, “A41.1”, “A41.59”, “B96.2”, “P35.2”, “A22.7”, “A28.0”, “A32.7”, “A39.4”, “A40.0”, “A40.3”, “A41.52”, “B37.7”, “I76”, “N39.0”, “R65.20”, “A20.7”, “A41”, “A41.02”, “A41.4”, “A41.51”, “A41.89”, “A41.9”, “B00.7”, “B95.61”, “B95.62”, “P36.0”, “P36.2”, “P36.8”, “R65.2”, “A28.2”, “A39.2”, “A41.0”, “A41.2”, “A41.5”, “A41.53”, “A41.8”, “A42.7”, “J44.0”, “P36.1”, “P36.3”, “P36.4”, “P36.9”, “P37.2”, “P37.5”, “A03.9”, “A04.7”, “A21.7”, “A23.9”, “A24.1”, “A39.3”, “A40.8”, “A40.9”, “A41.3”, “A41.50”, “A41.81”, “A54.86”, “B95.4”, “B95.6”, “J18.9”, “P36.5”, “R65.21” }
A02.1 - Salmonella sepsis
A26.7 - Erysipelothrix sepsis
A40 - Streptococcal sepsis
A40.1 - Sepsis due to streptococcus, group B
A41.01 - Sepsis due to Methicillin susceptible Staphylococcus aureus
A41.1 - Sepsis due to other specified staphylococcus
A41.59 - Other Gram-negative sepsis
B96.2 - Escherichia coli [E. coli] as the cause of diseases classified elsewhere
P35.2 - Congenital herpesviral [herpes simplex] infection
A22.7 - Anthrax sepsis
A28.0 - Pasteurellosis
A32.7 - Listerial sepsis
A39.4 - Meningococcemia, unspecified
A40.0 - Sepsis due to streptococcus, group A
A40.3 - Sepsis due to Streptococcus pneumoniae
A41.52 - Sepsis due to Pseudomonas
B37.7 - Candidal sepsis
I76 - Septic arterial embolism
N39.0 - Urinary tract infection, site not specified
R65.20 - Severe sepsis without septic shock
A20.7 - Septicemic plague
A41 - Other sepsis
A41.02 - Sepsis due to Methicillin resistant Staphylococcus aureus
A41.4 - Sepsis due to anaerobes
A41.51 - Sepsis due to Escherichia coli [E. coli]
A41.89 - Other specified sepsis
A41.9 - Sepsis, unspecified organism
B00.7 - Disseminated herpesviral disease
B95.61 - Methicillin susceptible Staphylococcus aureus infection as the cause of diseases classified elsewhere
B95.62 - Methicillin resistant Staphylococcus aureus infection as the cause of diseases classified elsewhere
P36.0 - Sepsis of newborn due to streptococcus, group B
P36.2 - Sepsis of newborn due to Staphylococcus aureus
P36.8 - Other bacterial sepsis of newborn
R65.2 - Severe sepsis
A28.2 - Extraintestinal yersiniosis
A39.2 - Acute meningococcemia
A41.0 - Sepsis due to Staphylococcus aureus
A41.2 - Sepsis due to unspecified staphylococcus
A41.5 - Sepsis due to other Gram-negative organisms
A41.53 - Sepsis due to Serratia
A41.8 - Other specified sepsis
A42.7 - Actinomycotic sepsis
J44.0 - Chronic obstructive pulmonary disease with acute lower respiratory infection
P36.1 - Sepsis of newborn due to other and unspecified streptococci
P36.3 - Sepsis of newborn due to other and unspecified staphylococci
P36.4 - Sepsis of newborn due to Escherichia coli
P36.9 - Bacterial sepsis of newborn, unspecified
P37.2 - Neonatal (disseminated) listeriosis
P37.5 - Neonatal candidiasis
A03.9 - Shigellosis, unspecified
A04.7 - Enterocolitis due to Clostridium difficile
A21.7 - Generalized tularemia
A23.9 - Brucellosis, unspecified
A24.1 - Acute and fulminating melioidosis
A39.3 - Chronic meningococcemia
A40.8 - Other streptococcal sepsis
A40.9 - Streptococcal sepsis, unspecified
A41.3 - Sepsis due to Hemophilus influenzae
A41.50 - Gram-negative sepsis, unspecified
A41.81 - Sepsis due to Enterococcus
A54.86 - Gonococcal sepsis
B95.4 - Other streptococcus as the cause of diseases classified elsewhere
B95.6 - Staphylococcus aureus as the cause of diseases classified elsewhere
J18.9 - Pneumonia, unspecified organism
P36.5 - Sepsis of newborn due to anaerobes
R65.21 - Severe sepsis with septic shock
Inpatient Indicator is any of:
Yes
The occurrence of Medical Claims with the following attributes:
Diagnosis Code, ICD-10 is any of: { “A02.1”, “A26.7”, “A40”, “A40.1”, “A41.01”, “A41.1”, “A41.59”, “B96.2”, “P35.2”, “A22.7”, “A28.0”, “A32.7”, “A39.4”, “A40.0”, “A40.3”, “A41.52”, “B37.7”, “I76”, “N39.0”, “R65.20”, “A20.7”, “A41”, “A41.02”, “A41.4”, “A41.51”, “A41.89”, “A41.9”, “B00.7”, “B95.61”, “B95.62”, “P36.0”, “P36.2”, “P36.8”, “R65.2”, “A28.2”, “A39.2”, “A41.0”, “A41.2”, “A41.5”, “A41.53”, “A41.8”, “A42.7”, “J44.0”, “P36.1”, “P36.3”, “P36.4”, “P36.9”, “P37.2”, “P37.5”, “A03.9”, “A04.7”, “A21.7”, “A23.9”, “A24.1”, “A39.3”, “A40.8”, “A40.9”, “A41.3”, “A41.50”, “A41.81”, “A54.86”, “B95.4”, “B95.6”, “J18.9”, “P36.5”, “R65.21” }
Suspected Inpatient Indicator is any of:
Yes
The occurrence of Medical Claims with the following attributes:
Diagnosis Code, ICD-10 is any of: { “A02.1”, “A26.7”, “A40”, “A40.1”, “A41.01”, “A41.1”, “A41.59”, “B96.2”, “P35.2”, “A22.7”, “A28.0”, “A32.7”, “A39.4”, “A40.0”, “A40.3”, “A41.52”, “B37.7”, “I76”, “N39.0”, “R65.20”, “A20.7”, “A41”, “A41.02”, “A41.4”, “A41.51”, “A41.89”, “A41.9”, “B00.7”, “B95.61”, “B95.62”, “P36.0”, “P36.2”, “P36.8”, “R65.2”, “A28.2”, “A39.2”, “A41.0”, “A41.2”, “A41.5”, “A41.53”, “A41.8”, “A42.7”, “J44.0”, “P36.1”, “P36.3”, “P36.4”, “P36.9”, “P37.2”, “P37.5”, “A03.9”, “A04.7”, “A21.7”, “A23.9”, “A24.1”, “A39.3”, “A40.8”, “A40.9”, “A41.3”, “A41.50”, “A41.81”, “A54.86”, “B95.4”, “B95.6”, “J18.9”, “P36.5”, “R65.21” }
Hospital Service Revenue Code is any of: { “0114”, “0115”, “0129”, “0142”, “0143”, “0144”, “0151”, “0156”, “0987”, “0100”, “0101”, “0118”, “0122”, “0127”, “0133”, “0152”, “0160”, “0117”, “0120”, “0130”, “0136”, “0140”, “0145”, “0150”, “0155”, “0158”, “0169”, “0012”, “0111”, “0113”, “0116”, “0119”, “0121”, “0125”, “0139”, “0141”, “0147”, “0148”, “0154”, “0157”, “0164”, “0167”, “0110”, “0112”, “0123”, “0124”, “0126”, “0128”, “0131”, “0132”, “0134”, “0135”, “0138”, “0153”, “0159” }
Hospitalization with COVID-19 + the occurrence of respiratory intubation and mechanical ventilation (procedure and diagnosis codes):
This measure was defined as:
1.1. The occurrence of Hospitalization with COVID
1.2. The occurrence of Respiratory intubation and mechanical ventilation (procedure and diagnosis codes), starting at least 0 days after the start of 1.1, and starting at most 30 days after the start of 1.1.
The event of interest in this group is 1.2.
The occurrence of respiratory intubation and mechanical ventilation (procedure and diagnosis codes):
The occurrence of Medical Claims with the following attributes:
Procedure Code, HCPCS and CPT is any of: { “31500”, “G8569”, “43753”, “94002”, “94657”, “94003”, “94004”, “94656” }
31500 - Intubation, endotracheal, emergency procedure
G8569 - PROLONGED POSTOPERATIVE INTUBATION (> 24 HRS) REQUIRED
43753 - Gastric intubation and aspiration(s) therapeutic, necessitating physician’s skill (eg, for gastrointestinal hemorrhage), including lavage if performed
94002 - Ventilation assist and management, initiation of pressure or volume preset ventilators for assisted or controlled breathing; hospital inpatient/observation, initial day
94657 - Ventilation assist and management, initiation of pressure or volume preset ventilators for assisted or controlled breathing; subsequent days
94003 - Ventilation assist and management, initiation of pressure or volume preset ventilators for assisted or controlled breathing; hospital inpatient/observation, each subsequent day
94004 - Ventilation assist and management, initiation of pressure or volume preset ventilators for assisted or controlled breathing; nursing facility, per day
94656 - Ventilation assist and management, initiation of pressure or volume preset ventilators for assisted or controlled breathing; first day
The occurrence of Medical Claims with the following attributes:
Diagnosis Code, ICD-10 is any of: { “Z99.1”, “Z99.11”, “Z99.12”, “J95.851”, “J95.85”, “J95.859”, “T88.4”, “T88.4XXA”, “T88.4XXS” }
Z99.1 - Dependence on respirator
Z99.11 - Dependence on respirator [ventilator] status
Z99.12 - Encounter for respirator [ventilator] dependence during power failure
J95.851 - Ventilator associated pneumonia
J95.85 - Complication of respirator [ventilator]
J95.859 - Other complication of respirator [ventilator]
T88.4 - Failed or difficult intubation
T88.4XXA - Failed or difficult intubation, initial encounter
T88.4XXS - Failed or difficult intubation, sequela
The occurrence of Medical Claims with the following attributes:
Procedure Code, ICD-10 is any of: { “09HN8BZ”, “0BH18EZ”, “0DH57BZ”, “0WHQ73Z”, “0WHQ7YZ”, “5A09457”, “09HN7BZ”, “5A09557”, “0CHY7BZ”, “5A09357”, “5A1945Z”, “5A1955Z”, “0BH17EZ”, “0BH13EZ”, “0CHY8BZ”, “0DH58BZ”, “5A1935Z” }
09HN8BZ - Insertion of Airway into Nasopharynx, Via Natural or Artificial Opening Endoscopic
0BH18EZ - Insertion of Endotracheal Airway into Trachea, Via Natural or Artificial Opening Endoscopic
0DH57BZ - Insertion of Airway into Esophagus, Via Natural or Artificial Opening
0WHQ73Z - Insertion of Infusion Device into Respiratory Tract, Via Natural or Artificial Opening
0WHQ7YZ - Insertion of Other Device into Respiratory Tract, Via Natural or Artificial Opening
5A09457 - Assistance with Respiratory Ventilation, 24-96 Consecutive Hours, Continuous Positive Airway Pressure
09HN7BZ - Insertion of Airway into Nasopharynx, Via Natural or Artificial Opening
5A09557 - Assistance with Respiratory Ventilation, Greater than 96 Consecutive Hours, Continuous Positive Airway Pressure
0CHY7BZ - Insertion of Airway into Mouth and Throat, Via Natural or Artificial Opening
5A09357 - Assistance with Respiratory Ventilation, Less than 24 Consecutive Hours, Continuous Positive Airway Pressure
5A1945Z - Respiratory Ventilation, 24-96 Consecutive Hours
5A1955Z - Respiratory Ventilation, Greater than 96 Consecutive Hours
0BH17EZ - Insertion of Endotracheal Airway into Trachea, Via Natural or Artificial Opening
0BH13EZ - Insertion of Endotracheal Airway into Trachea, Percutaneous Approach
0CHY8BZ - Insertion of Airway into Mouth and Throat, Via Natural or Artificial Opening Endoscopic
0DH58BZ - Insertion of Airway into Esophagus, Via Natural or Artificial Opening Endoscopic
5A1935Z - Respiratory Ventilation, Less than 24 Consecutive Hours
Footnotes
Funding/Support: No further funding was received for this study.
Financial Disclosure: All authors have completed the Unified Competing Interest form (available on request from the corresponding author and declare:
Dr. Schneeweiss M. has no relevant financial interests to report. Ms. Leonard is employee of HealthVerity Inc. and holds equity
Dr. Schneeweiss S. is the principal investigator of investigator-initiated grants to the Brigham and Women’s Hospital from the FDA, NIH, PCORI, Bayer, Vertex, and Boehringer Ingelheim unrelated to the topic of this study. He is a consultant to Aetion, a software manufacturer of which he owns equity. His interests were declared, reviewed, and approved by the Brigham and Women’s Hospital and Partners HealthCare System in accordance with their institutional compliance policies.
Drs. Weckstein and Rassen are employees of Aetion Inc. and hold equity.
References
