Sociodemographic Characteristics, Comorbidities, and Mortality Among Persons Diagnosed With Tuberculosis and COVID-19 in Close Succession in California, 2020

IMPORTANCE Tuberculosis (TB) and COVID-19 are respiratory diseases that disproportionately occur among medically underserved populations; little is known about their epidemiologic intersection. OBJECTIVE To characterize persons diagnosed with TB and COVID-19 in California. DESIGN, SETTING, AND PARTICIPANTS This cross-sectional analysis of population-based public health surveillance data assessed the sociodemographic, clinical, and epidemiologic characteristics of California residents who were diagnosed with TB (including cases diagnosed and reported between September 3, 2019, and December 31, 2020) and COVID-19 (including confirmed cases based on positive results on polymerase chain reaction tests and probable cases based on positive results on antigen assays reported through February 2, 2021) in close succession compared with those who were diagnosed with TB before the COVID-19 pandemic (between January 1, 2017, and December 31, 2019) or diagnosed with COVID-19 alone (through February 2, 2021). This analysis included 3 402 713 California residents with COVID-19 alone, 6280 with TB before the pandemic, and 91 with confirmed or probable COVID-19 diagnosed within 120 days of a TB diagnosis (ie, TB/COVID-19). EXPOSURES Sociodemographic characteristics, medical risk factors, factors associated with TB severity, and health equity index. MAIN OUTCOMES AND MEASURES Frequency of reported successive TB and COVID-19 (TB/COVID-19) diagnoses within 120 days, frequency of deaths, and age-adjusted mortality rates. RESULTS Among the 91 persons with TB/COVID-19, the median age was 58.0 years (range, 3.0-95.0 years; IQR, 41.0-73.0 years); 52 persons (57.1%) were male; 81 (89.0%) were born outside the US; and 28 (30.8%) were Asian or Pacific Islander, 4 (4.4%) were Black, 55 (60.4%) were Hispanic or Latino, 4 (4.4%) were White. The frequency of reported COVID-19 among those who received a TB diagnosis between September 3, 2019, and December 31, 2020, was 225 of 2210 persons (10.2%), which was similar to that of the general population (3 402 804 of 39 538 223 persons [8.6%]). Compared with persons with TB before the pandemic, those with TB/COVID-19 were more likely to be Hispanic or Latino (2285 of 6279 persons [36.4%; 95% CI, 35.2%-37.6%] vs 55 of 91 persons [60.4%; 95% CI, 49.6%-70.5%], respectively; P < .001), reside in low health equity census tracts (1984 of 6027 persons [32.9%; 95% CI, 31.7%-34.1%] vs 40 of 89 persons [44.9%; 95% CI, 34.4%55.9%]; P = .003), live in the US longer before receiving a TB diagnosis (median, 19.7 years [IQR, 7.2-32.3 years] vs 23.1 years [IQR, 15.2-31.5 years]; P = .03), and have diabetes (1734 of 6280 persons [27.6%; 95% CI, 26.5%-28.7%] vs 42 of 91 persons [46.2%; 95% CI, 35.6%-56.9%]; P < .001). The frequency of deaths among those with TB/COVID-19 successively diagnosed within 30 days (8 of 34 persons [23.5%; 95% CI, 10.8%-41.2%]) was more than twice that of persons with TB before the (continued) Key Points Question What are the sociodemographic, clinical, and epidemiologic characteristics of persons diagnosed with tuberculosis (TB) and COVID-19 in close succession in California? Findings In this cross-sectional analysis of public health surveillance records from California residents, 91 individuals diagnosed with TB and COVID-19 more commonly had Hispanic or Latino ethnicity, diabetes, and residence in a low health equity census tract compared with those who received a TB diagnosis before the COVID-19 pandemic. Mortality rates among those diagnosed with TB and COVID-19 in close succession were higher than mortality rates among those with TB before the COVID-19 pandemic and those with COVID19 alone. Meaning The findings of this analysis suggest that addressing long-standing health inequities and integrating prevention measures for COVID-19 and TB in California may reduce the co-occurrence of these diseases and prevent deaths. Author affiliations and article information are listed at the end of this article. Open Access. This is an open access article distributed under the terms of the CC-BY License. JAMA Network Open. 2021;4(12):e2136853. doi:10.1001/jamanetworkopen.2021.36853 (Reprinted) December 3, 2021 1/9 Downloaded From: https://jamanetwork.com/ on 12/08/2021 Abstract (continued)continued) pandemic (631 of 5545 persons [11.4%; 95% CI, 10.6%-12.2%]; P = .05) and 20 times that of persons with COVID-19 alone (42 171 of 3 402 713 persons [1.2%; 95% CI, 1.2%-1.3%]; P < .001). Persons with TB/COVID-19 who died were older (median, 81.0 years; IQR, 75.0-85.0 years) than those who survived (median, 54.0 years; IQR, 37.5-68.5 years; P < .001). The age-adjusted mortality rate remained higher among persons with TB/COVID-19 (74.2 deaths per 1000 persons; 95% CI, 26.2122.1 deaths per 1000 persons) compared with either disease alone (TB before the pandemic: 56.3 deaths per 1000 persons [95% CI, 51.2-61.4 deaths per 1000 persons]; COVID-19 only: 17.1 deaths per 1000 persons [95% CI, 16.9-17.2 deaths per 1000 persons]). CONCLUSIONS AND RELEVANCE In this cross-sectional analysis, TB/COVID-19 was disproportionately diagnosed among California residents who were Hispanic or Latino, had diabetes, or were living in low health equity census tracts. These results suggest that tuberculosis and COVID-19 occurring together may be associated with increases in mortality compared with either disease alone, especially among older adults. Addressing health inequities and integrating prevention efforts could avert the occurrence of concurrent COVID-19 and TB and potentially reduce deaths. JAMA Network Open. 2021;4(12):e2136853. doi:10.1001/jamanetworkopen.2021.36853


Introduction
The US has a low tuberculosis (TB) burden, and California, a diverse state with 40 million residents, reports one-quarter of the nation's cases. 1 Travelers to California were among the first to receive COVID-19 diagnoses in the US, producing subsequent widespread community transmission. 2 Although TB and COVID-19 share some medical risk factors, little is known about the epidemiologic intersection of these primarily respiratory diseases, especially in settings with low TB incidence. We aimed to describe the epidemiologic characteristics and deaths associated with these diseases among California residents who were diagnosed with TB and COVID-19 in close succession in comparison with those among individuals diagnosed with TB before the pandemic.

Methods
This cross-sectional analysis was approved by the California Department of Public Health and the Centers for Disease Control and Prevention. Because public health surveillance data were used for emergency response purposes, a nonresearch determination was made by both institutions, and informed consent was not required. The analysis followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Diagnoses of persons with both TB and COVID-19 (TB/COVID-19) were defined using 2 criteria: (1) a diagnosis of TB and/or COVID-19 in California in 2020 and (2) successive diagnoses of TB and COVID-19 within 120 days. We selected the 120-day interval to represent persons who had acute illness with TB and COVID-19 in close succession. Although symptomatic COVID-19 has rapid onset and is typically diagnosed within days, TB generally has a slower onset, and diagnosis may take several weeks to months.
We first identified California residents with active TB who received a diagnosis between September 3, 2019 (ie, within 120 days of January 1, 2020), and December 31, 2020 (Figure 1). We then cross-matched these records with all confirmed (defined as a positive result on a polymerase chain reaction test) or probable (defined a positive result on an antigen assay) COVID-19 cases reported through February 2, 2021, the cutoff date when the match with TB cases was performed.
We used surveillance records from the California Reportable Disease Information Exchange (CalREDIE), the communicable disease surveillance database maintained by the California Department of Public Health. We used a probabilistic algorithm to match first name, last name, date of birth, sex, and zip code in Match*Pro software, version 1.6.2 (National Cancer Institute), and required an exact match on at least one of first name, last name, or date of birth.
The TB diagnosis date was the earliest recorded among the report date, treatment start date, or specimen collection date during which a positive result on a blood culture or nucleic acid amplification analysis was recorded. We defined the COVID-19 diagnosis date as the earliest among specimen collection dates during which a positive test result was recorded. We considered positive results for cavitary disease (via chest radiography or computed tomography) and sputum smear tests for acid-fast bacilli to potentially indicate advanced pulmonary TB. Disseminated TB was defined as a positive blood culture result for Mycobacterium tuberculosis complex, meningeal involvement, miliary disease, or both extrapulmonary and pulmonary disease.
For persons diagnosed with TB before the COVID-19 pandemic (January 1, 2017, to December 31, 2019), we used TB surveillance records to ascertain deaths before or during TB treatment. In general, 174 TB cases were diagnosed per month from January 1, 2017, to December 31, 2019. From September 3, 2019, to December 31, 2020, approximately 147 TB cases were diagnosed per month.
For persons with TB/COVID-19 and persons with COVID-19 alone, we ascertained death status as of April 10, 2021, by using COVID-19 surveillance records and by cross-referencing reported COVID-19 cases in the California vital statistics database. Cases were assigned to quartiles according to scores on the California Healthy Places Index, 3 a multidomain composite social inequity index, by census tract of residence (with quartile 1 indicating least advantaged and quartile 4 indicating most advantaged).

Statistical Analysis
Matched persons with TB/COVID-19 were compared with all 6280 persons diagnosed with TB before the COVID-19 pandemic, with frequency of death included in comparisons. Using the same numerators, we directly age adjusted mortality rates per 1000 residents to the 2020 California Department of Finance standard population (10-year age strata flanked by the age groups of Յ14 years and Ն75 years). 4 One person was diagnosed with TB in December 2019 and with COVID-19 in April 2020; this person was part of the TB/COVID-19 analysis group and excluded from the prepandemic TB analysis group. For statistical comparisons, we used 2-sided χ2 or Fisher exact tests for categorical data and Wilcoxon 2-sample tests for continuous data. Log-normal CIs for rate ratios were calculated using SAS software, version 9.4 (SAS Institute, Inc). The significance threshold was P = .05.

Discussion
This cross-sectional analysis is, to our knowledge, one of the first population-based analyses of TB and COVID-19 surveillance data in a low-incidence setting for TB. We found that California residents with TB/COVID-19 had characteristics that were generally similar to those of persons with TB before the COVID-19 pandemic.

Limitations
This analysis has limitations. We were unable to measure certain clinical factors associated with disease severity, such as hospitalization, or evaluate the circumstances associated with deaths.
Whereas TB surveillance data were generally complete, a high frequency of missing sociodemographic data in the COVID-19 surveillance database precluded direct case-level comparisons between persons with TB/COVID-19 and persons with COVID-19 alone in 2020. The small number of persons with TB/COVID-19 produced wide ranges in CIs for rate comparisons.
Nevertheless, we likely underestimated mortality among persons with TB/COVID-19 because follow-up surveillance reporting for TB, which captures death at any point during TB treatment, is not yet complete for 2020. In addition, this lack of follow-up data did not allow comparison of outcomes among persons diagnosed with TB in 2020 who did not have COVID-19. Additional reasons we may have underestimated mortality include the possibility that persons who survived their initial illness could have survived for up to 120 days to allow for the second diagnosis, and COVID-19-associated deaths peaked in California after our observation period. 5 Underdetection of both COVID-19 and TB may have occurred in 2020, and the consequences this underdetection may have had for our surveillance results are an important area of future inquiry.

Conclusions
This cross-sectional analysis found that during the first year of the COVID-19 pandemic, California residents with TB/COVID-19 had higher mortality than those with either disease alone; however, additional studies in the US are needed to assess the generalizability of these findings. Tuberculosis disproportionately occurs in medically and socially vulnerable communities, and these results suggest potential benefit from the integration of TB and COVID-19 prevention efforts, such as combining COVID-19 vaccination outreach with targeted screening for TB. Reductions in reported TB during the pandemic likely reflected, in part, decreased TB detection in the US. 1 These reductions highlight the need for health care professionals to consider TB as a potential diagnosis among persons at risk and in the appropriate clinical context, including the possibility of M tuberculosis coinfection among persons with positive test results for SARS-CoV-2.