Abstract
Objective:
Data were essential to public health decision-making during the COVID-19 pandemic, yet no single data source was adequate for Tribes in Montana and Wyoming. We outlined data access, availability, and limitations for COVID-19 pandemic surveillance response to improve future data exchange.
Materials and Methods:
The Rocky Mountain Tribal Epidemiology Center (RMTEC) used various data sources to deliver data on the number of COVID-19 cases, deaths, and vaccinations at local, state, and regional levels to inform Tribes in Montana and Wyoming. RMTEC reviewed state, federal, and public datasets and then attached a score to each dataset for completeness of demographic information, including race, geographic level, and refresh rate.
Results:
The RMTEC COVID-19 response team shared data weekly on the number of COVID-19 cases, deaths, and vaccinations distributed and the percentage of the population vaccinated with Tribal health departments in Montana and Wyoming. The Indian Health Service Epidemiology Data Mart dataset scored the highest (24 of 30), followed by datasets from Montana (18 of 30) and Wyoming (22 of 30). Publicly available datasets scored low largely due to data aggregation across larger geographic areas and lack of demographic variables.
Practice Implications:
The absence of data on race and ethnicity from publicly available data and lack of access to real-time data limited RMTEC’s ability to provide Tribal-specific updates on COVID-19 cases, deaths, and vaccinations to Tribal health departments. RMTEC should be fully funded to provide the necessary resources for data management and the capacity to respond to data requests from Tribal health departments and their programs to address current and future pandemics. Federal and state agencies should also be educated on Tribal Epidemiology Centers’ public health authority status to improve access to infectious disease data among those agencies.
The COVID-19 pandemic, caused by the SARS-CoV-2 virus, has killed more than 5 million people 1 worldwide since December 2019. 2 Available federal and state data show that American Indian and Alaska Native (AI/AN) people are disproportionately affected by COVID-19 diagnoses and death.3-6 The COVID-19 case fatality rate per 1000 COVID-19 cases among AI/AN people (29.4) was 1.7 times higher than the rate among non-Hispanic White people (17.0) from March 13 through November 30, 2020. 3
The Rocky Mountain Tribal Epidemiology Center (RMTEC), a division of the Rocky Mountain Tribal Leaders Council in Billings, Montana, is 1 of 12 Tribal Epidemiology Centers (TECs) established under the Indian Health Care Improvement Act 7 as a public health authority and serves more than 70 000 AI/AN people in Montana and Wyoming. In Montana and Wyoming, there are 6 Indian Health Service (IHS)–operated service units, 2 Tribally operated health departments, and 5 Urban Indian Health Organizations. RMTEC’s mission is to empower Tribes with reliable data for public health planning and program development. RMTEC’s protocol is congruent with Tribal data sovereignty and the inherent understanding that Tribal data belong to the Tribes first and foremost. In response to the COVID-19 pandemic, a multidisciplinary team was formed to acquire and report surveillance data to guide recommendations and inform Tribal public health response in Montana and Wyoming.
The biggest challenge RMTEC experienced during the COVID-19 pandemic was a lack of access to real-time data from federal agencies and state departments of health. While memorandums of understanding and data-sharing agreements were established, they were not necessarily fulfilled promptly, which affected competent surveillance. In addition, issues of duplicative surveillance systems and missing key trends on cases or vaccinations limited informed public health decision-making that is vital to mitigation. To circumvent these challenges, RMTEC sought data from Centers for Disease Control and Prevention (CDC) case investigation data, IHS, the states, and other institutions to generate numerical and visual reports for regional Tribal health directors. We describe the methods used by RMTEC during the COVID-19 surveillance response developed from April 2020 through December 2021 for Tribes in Montana and Wyoming.
Methods and Materials
The RMTEC COVID-19 workgroup met weekly via virtual meeting platforms beginning April 2020 to discuss and select COVID-19 indicators for regional Tribal health directors’ weekly and biweekly (every other week) reports. The multidisciplinary team consisted of 3 epidemiologists, a training coordinator, a communication specialist, a statistician, a project director, and the RMTEC director. During that time, the workgroup identified the following data sources and measures:
CDC publicly available data: National-level COVID-19 case data were made available by CDC on January 21, 2020, 8 and vaccine distribution data were added as the first COVID-19 vaccines (Pfizer-BioNTech and Moderna mRNA) became available in February 2021. This dataset includes data on IHS-provided doses available through the Vaccine Administration Management System, 6 an online tool used by CDC and IHS to manage vaccine administration and tracking in public health clinics. From this dataset, RMTEC collected data on case numbers and vaccination rates at the national, state, and county level on a weekly basis. The biggest limitation was the exclusion of data on race and ethnicity in vaccine administration.
CDC Data Collation and Integration for Public Health Event Response (DCIPHER) 7 : In the first few months of the pandemic, IHS established a data-sharing agreement with CDC to securely receive privately protected data through cloud-based HHS Protect, which allowed the TECs access to these data. 8 CDC DCIPHER is part of a coordinated COVID-19 response in which local jurisdictions share deidentified patient-level data for each case with the CDC Case Surveillance Task Force and the Surveillance Review and Response Group. DCIPHER contains deidentified COVID-19 case data, including patient demographic characteristics, exposure history, disease severity indicators and outcomes, clinical data, laboratory diagnostic test results, and comorbidities. RMTEC analyzed these data for dissemination to Tribal health directors using SAS version 9.4 (SAS Institute, Inc) to produce biweekly (every other week) reports that contained (1) COVID-19 diagnosis as laboratory confirmed or probable, (2) whether the patient was symptomatic or asymptomatic, (3) the patient’s biological sex and pregnancy status, and (4) the outcomes of hospitalization intensive care unit (ICU) admission, (5) death, and (6) laboratory-confirmed cases by month.
IHS publicly available data: The IHS Billings Area Office (BAO-IHS) COVID-19 information web page 9 provides data on testing and vaccine distribution for each IHS area. IHS provides measures for total COVID-19 vaccinations: the number of doses distributed and doses administered. Vaccine administration and distribution are limited to facilities that chose the IHS authority for vaccine distribution. Alternatively, Tribal jurisdictions may have chosen to administer doses in collaboration with their state. RMTEC selected the following measures from the IHS web page: vaccine “doses distributed,” “doses administered,” and COVID-19 cases “tested,” “positive,” and “negative.” RMTEC calculated the number of cumulative tests and pending tests to report incident COVID-19 cases, testing outcomes (positive, negative, or pending), delivered vaccine doses, and administered doses in the BAO-IHS.
IHS Epidemiology Data Mart (EDM) 10 : Prior to COVID-19, RMTEC received previous fiscal year deidentified IHS service unit data once annually related to encounters in health care facilities, which included information on the location of treatment, clinic, health care provider, medications, diagnosis codes, and others. Using the EDM, RMTEC has developed multiple reports, including annual and quarterly Tribe-specific reports on health outcomes for Tribal health directors and COVID-19 and comorbidity reports that were disseminated to all Tribal health directors in Montana and Wyoming. In addition, RMTEC used the EDM to analyze encounter data from September 1, 2020, through April 1, 2021, using SAS version 9.4 (SAS Institute, Inc) (Table 1). The IHS EDM was queried for International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes 11 for diagnoses of known, probable, or suspected COVID-19 cases with and without comorbidities of interest. Diagnoses counts were restricted to Countable Active Indian Registrants (CAIR) visiting IHS Billings Area Office service units from April 1, 2020, through September 1, 2021. All encounters (patients, outpatients, and telemedicine) were included. Comorbidities were selected for analysis by pairing the most frequent comorbidities in raw COVID-19 publicly available case data 12 for all races to IHS-identified health disparities affecting AI/AN people. 13 Comorbidity diagnoses must have been made or referred to within the past 3 years (any encounter after May 31, 2017), with the exception of diagnoses of influenza and pneumonia, which must have occurred in the 3 months before a diagnosis of COVID-19. Unlike chronic conditions, influenza and pneumonia cause short-term changes in the immune system that may alter the severity of COVID-19.14-16 Therefore, the search for these diagnoses was limited to 3 months prior to the diagnosis of COVID-19. For unclear descriptions of comorbid conditions in federal references,12,13 the more precise definition was used for the analysis. For example, the CDC-identified comorbidity “heart failure” was used for analysis instead of the less precise IHS-recognized disparity “diseases of the heart.” RMTEC did not differentiate between outcomes and coexisting conditions.
Montana and Wyoming department of health web pages: Montana17,18 and Wyoming 19 state data on case counts and vaccinations per county are made available on each workday with occasional delays, particularly after weekends and holidays. RMTEC collected and summed weekly county-level data on case and vaccine numbers for counties within the reservation Contract Health Service Delivery Area (CHSDA). CHSDAs are IHS-recognized geographic boundaries of a county or counties “which include all or part of a reservation and any county or counties which have a common boundary with the reservation.” 20 Where CHSDAs include more than 1 reservation, RMTEC used percentages of the population agreed upon by the affected Tribes to attribute the counts proportionally. These data do not have a race variable. For rates and percentages, RMTEC used Surveillance, Epidemiology, and End Result (SEER) 21 county-level census data as the denominator.
The New York Times: In addition to the aforementioned data sources, RMTEC sought information from other available sources to develop a dashboard of indicators: COVID-19 cases, death due to COVID-19, and COVID-19 vaccine administration. The New York Times COVID-19 cases web page 22 provides county-level data on cases, hospitalizations, and vaccinations. These data do not include a race variable.
Comorbidities associated with COVID-19 infection a that impacted American Indian and Alaska Native people in the Billings Area of Montana and Wyoming, April 1, 2020, through September 1, 2021
Abbreviation: ICD-10-CM, International Classification of Diseases, Tenth Revision, Clinical Modification.
Comorbidities were identified through analysis of Indian Health Service Epidemiology Data Mart (EDM) data. 10
Data source: National Center for Health Statistics. 11
Chronic lower respiratory diseases include chronic obstructive pulmonary disease, chronic bronchitis, emphysema, and asthma.
Cerebrovascular diseases are conditions affecting blood flow and blood vessels in the brain, such as stenosis, thrombosis, embolism, or hemorrhage.
RMTEC evaluated multiple data sources7-10,17-19,22 for completeness and availability, creating Tribal- and regional-specific COVID-19 data visualizations for Tribal health departments in the Billings area (Montana and Wyoming). These data sources provided variables including number of cases, deaths, hospitalizations, and vaccinations at varying levels of detail and geography.
RMTEC scored these datasets7-10,17-19,22 and 2 additional datasets (Montana and Wyoming departments of public health data by data-sharing agreement) on their refresh rate, race variable selection, geographic level, and demographic level for both case and death data. Data refresh indicates when data were available to RMTEC: a score of 3 indicates data that were updated daily, 2 indicates weekly, 1 indicates monthly, and 0 indicates that data were released infrequently or had to be requested. Data on race selection (using 5 races as defined by US Office of Management and Budget standards) were scored 3 when comprehensive demographic data, including combined race, age, sex, and comorbidities, were available; 2 if combination race data were available; 1 if single-race data only were available; and 0 if race data were not available. The geographic level of the data was scored 3 when data granularity was at the county level, 2 when at the state level, 1 when at the region level, and 0 when at the national level only. Race completeness was scored 3 when 95% of cases or deaths had race data, 2 when 75% did, 3 when 25% did, and 0 when no race data were available. The highest possible score was 30.
Results
The datasets available to RMTEC in November 2020 for use in county-level data analysis and visualizations for AI/AN populations in the Billings area varied in their available data variables. Federal publicly available datasets 9 and news organizations 22 lacked data on race; federal datasets accessible to public health authorities7,10 contained data on race when provided by the local reporting entity, and publicly available data from state departments of health17-19 included data on race at the state level but not the county level (Table 2).
Data sources and variables used to create Tribal- and regional-specific COVID-19 data visualizations for American Indian and Alaska Native people in the Billings Area of Montana and Wyoming, April 2020–February 2022
Abbreviations: CDC, Centers for Disease Control and Prevention; CHSDA, Contract Health Service Delivery Area; ICU, intensive care unit; IHS, Indian Health Service; RMTEC, Rocky Mountain Tribal Epidemiology Center.
RMTEC found a lack of data on race, missing variables, and infrequent refresh rates in these datasets. When scoring datasets, we found that the IHS EDM dataset 10 scored the highest overall (24 of a possible score of 30) (Table 3). This high score was due to the completeness of demographic, geographic, and race data in cases and deaths; however, this dataset was refreshed only annually (0 of a possible score of 3 for both case and death data). Montana and Wyoming state department of health–protected datasets available by data-sharing agreement scored the next highest (18 and 22, respectively, of a possible score of 30). Publicly available web pages and dashboards8,9,17,18,22 scored lower because the data were aggregated for larger geographic areas, lacked demographic and race data, and were updated infrequently. The lowest score was the IHS COVID-19 web page (4 of a possible score of 30).
COVID-19 datasets available to RMTEC a in the first 6 months of the pandemic for use in Contract Health Service Delivery Area–level data analysis, scored on refresh rate, availability of race data, and demographic and geographic specificity
Abbreviations: CDC, Centers for Disease Control and Prevention; IHS, Indian Health Service; RMTEC, Rocky Mountain Tribal Epidemiology Center.
Datasets accessed November 2020. Data were scored on a scale from 1 to 3, with 3 indicating better data quality.
Data source: Centers for Disease Control and Prevention. 5
Data source: Lee et al. 7
Data source: Indian Health Service. 9
Data source: Indian Health Service. 10
Data source: Montana.gov. 17
Data source: Wyoming Department of Health. 19
Data source: The New York Times. 22
Data refresh indicates when data were available to RMTEC: a score of 3 indicates data that were updated daily, 2 for weekly, 1 for monthly, and 0 when data were released infrequently or had to be requested.
Data on race selection (using 5 races as defined by US Office of Management and Budget standards) were scored 3 when comprehensive demographic data, including combined race, age, sex, and comorbidities, were available; 2 if combination race data were available; 1 if single race data only were available; and 0 if race data were not available.
The geographic level of the data was scored 3 when data granularity was at the county level, 2 at the state level, 1 at the region level, and 0 at the national level only.
Race completeness was scored 3 when 95% of cases or deaths had race data, 2 when 75% did, 3 when 25% did, and 0 when no race data were available. CDC caps race completeness at 95%.
Of a possible score of 30.
RMTEC began developing and disseminating COVID-19–specific reports to Tribal health directors in Montana and Wyoming in April 2020 using publicly available data. The incompleteness of data sources in November 2020 necessitated that data be pulled from multiple sources. The following reports and data visualizations used CHSDA boundaries to define Tribal nations.
Biweekly COVID-19 Case Report
Using CDC DCIPHER data, RMTEC has developed biweekly Tribe-specific updates since October 2020. These reports include frequencies and data visualization of percentages.
Weekly Report
RMTEC has developed weekly aggregate updates since April 2020 using data acquired from CDC, BAO-IHS, the Montana Department of Public Health and Human Services, and the Wyoming Department of Health. RMTEC created county-level data visualizations for the percentage of the eligible population who is fully vaccinated, number of vaccinations by dose, total number of cases, and number of new and active cases.
COVID-19 Dashboard
RMTEC created a daily COVID-19 dashboard using Tableau version 2020.4 (Salesforce, Inc) to show a summary of COVID-19 outcomes and vaccination progress for Montana and Wyoming. The dashboard shows new and cumulative COVID-19 data at the state and county levels.
IHS EDM Report
An aggregate report was generated, using encounter data from September 2020 through April 2021, and made available on September 3, 2021. Results were stratified by age and sex and included crude rates to allow comparison with state and national data.
Discussion
The COVID-19 pandemic uncovered unique challenges in all public health systems, especially those serving AI/AN populations. Datasets have highlighted the COVID-19 impact nationally, but the lack of datasets with race and ethnicity variables can lead to a lack of timely surveillance data. To mitigate these shortcomings often found in state and national datasets, RMTEC sought publicly available datasets to improve the Tribal response to COVID-19. Data on race, when present, are notoriously inaccurate and can minimize and obscure disparities.23,24 A 2015 study on hospital discharge data found that 55.4% of discharge records of individuals matching Tribal enrollment records were missing race or were miscoded as “White.” 25
Despite data availability and access challenges, RMTEC increased surveillance reporting with weekly and biweekly (every other week) updates and data visualization. RMTEC faced challenges in analyses attributable to using external data sources and having no influence in the variables available or the length of interval between collection and distribution.
Montana and Wyoming supplied county-level data. Data on race are not reported by county; therefore, measures of racial disparity cannot be assessed. The data are not updated on weekends and holidays. Also, the data lack information on hospitalizations and recoveries due to COVID-19.
The missing variables of race and ethnicity in these datasets make it difficult to show the impact of the pandemic on each Tribe and to implement activities based on accurate data. Inequities in public health surveillance systems may not identify AI/AN populations, which can perpetuate the continued invisibility of AI/AN populations through data erasure. The lack of data collection exacerbates health disparities because data are directly tied to policy decisions and resource allocation.26,27 In addition to inadequate resource allocation, the lack of quality and complete data obscures the incidence of disease among AI/AN populations, allowing for the possibility of health disparities to widen because of misinformed policies and programs.
Practice Implications
RMTEC established timely surveillance for the COVID-19 pandemic to provide reliable data to Tribes in Montana and Wyoming. From April 2020 through December 2021, the surveillance strengthened the overall Tribal response to identify cases for improved coordination and surveillance capacity in Tribal health departments and RMTEC. For example, a member Tribe requested that RMTEC evaluate whether the Tribe should relax its COVID-19 precautions when the pandemic was in a downturn. Rather than give a short answer, RMTEC used a CDC decision matrix and data from the sources listed herein to create a heat map that ranked indicators as favorable, neutral, or not favorable for reopening so the Tribe could make a decision based on the indicators most important to them.
Epidemiological surveillance is critical to prevent the spread of COVID-19. Readily available information resources usable by the public can help promote interoperability and openness, which improves the health and safety of all. 28 Data initiatives that promote data sovereignty through increased sharing of data collection, enhanced data analysis, and designing data-driven solutions are important for transparency and understanding the impact of COVID-19. 29
RMTEC has relied on data to understand the effects of the COVID-19 pandemic on AI/AN communities. Incomplete data and varied reporting standards directly affect the ability of AI/AN populations to address the COVID-19 pandemic. It is important that each level of government address limitations in the quality, quantity, and access to data on AI/AN populations. AI/AN populations deserve timely, relevant, high-quality data to inform their own pandemic response. An article by Conger et al 26 identifying data as a strategic resource for Indigenous nations recognized that the external control of data means that data do not exist to inform Tribal needs, and existing data cannot be aggregated in ways that are meaningful to Tribes.
The COVID-19 case surveillance system is passive; therefore, the data underestimate the true number of cases through underdiagnosis or underreporting. Completeness of reporting is influenced by multiple factors, including the availability of diagnostic testing, the availability of resources, and the priorities of public health jurisdictions. In addition, reporting is voluntary and varies by state. Differences exist between state-specific databases and data from CDC’s COVID-19 surveillance database. The case report form captures several outcomes, including hospitalization, ICU admission, and death. When outcomes are not known at the time of reporting, they are classified as “unknown” or “missing.”
RMTEC, like many public health authorities, had to decide between providing incomplete and possibly incorrect data quickly and providing data that, while corrected, were no longer contemporary. RMTEC chose to provide the latest available data because the data were more useful to Tribes in Montana and Wyoming than data that may be corrected after an unspecified delay. One challenge of public health interventions is that the agencies generating the clinical data are not necessarily responsible for making appropriate interventions. Interventions can be ineffective when there is a lag or miscommunication in data. In the interest of effective response, public health agencies at all levels must be capable of sharing data as completely and quickly as possible while maintaining patient privacy and Tribal data sovereignty.
For Tribal nations, the responsibilities for translating public health data to interventions are held by various federal, state, and Tribal entities (Figure), which is unlike many state and local health departments, in which all elements of the intervention are held by a single organization. This disjointed approach to public health was exposed during the COVID-19 pandemic; many Tribes lacked access to their own data, which limited interventions to address the pandemic. The data provided by RMTEC allow Tribal health departments to implement appropriate public health interventions, including school closures and promoting vaccine uptake in their communities. Moving forward, several key components must be effectively executed, the most important of which is fully funding RMTEC to provide the necessary resources for data management and the capacity to secure the data requests for both RMTEC and Tribal health departments. Federal and state agencies should give RMTEC access to infectious disease data. Data requests from RMTEC must be filled in a period that allows for an actionable intervention based on the incubation period of the infectious disease. Early in the pandemic, a bicameral, bipartisan group of lawmakers wrote to the then–United States Secretary of Health Azar and CDC Director Redfield to warn that the US Department of Health and Human Services (HHS) refusals to share pandemic data with TECs appeared to violate federal law and hampered the efforts of AI/AN communities to respond. 30 In June 2021, the US House of Representatives passed the Tribal Health Data Improvement Act of 2020, which directs HHS to expand Tribal access to public health data. 31 A March 2022 Government Accountability Office report 32 examined the effectiveness of TECs and found that existing HHS policies or lack of policies inhibited TECs from fulfilling their core functions. The report applies to any data sharing, but because investigation occurred during the COVID-19 pandemic, the message was clear that the need for solutions is urgent. The Government Accountability Office report reiterated that, according to federal law, HHS must give TECs access to protected health data and recommended that HHS institute policies, guidance, and procedures to ensure timely data access. The recommendations, in brief, are that HHS, IHS, and CDC (1) establish policies clarifying datasets, which are to be made available to TECs, and (2) create policies on how TECs can request such data, how the requests will be reviewed by the agency, and the time frame in which the agency will respond. The report describes an improved framework for federal agencies to share data with TECs, but it could also be applied to state agencies.
Proposed improved data flow from federal and state entities to Rocky Mountain Tribal Epidemiology Center (RMTEC) and Tribal public health departments for faster and more effective public health policy development and timely interventions.
Despite the difficulties of accessing real-time data as described in this study, some federal and state agencies have partnered successfully with TECs. For example, Montana and RMTEC partnered to study the racial discrepancy of COVID-19 incidence and mortality in Montana and revealed a case fatality rate among AI/AN people that was 1.7 times higher than among White people. 3
We are confident that the partnership we share with federal and state agencies will continue to improve, particularly if the steps laid out by the Government Accountability Office are made. We look forward to many successful endeavors with our partners to the benefit of our member Tribes and all people.
Footnotes
Acknowledgements
The authors give special thanks to the Montana Department of Public Health and Human Services, the Wyoming Department of Health, CDC, and IHS for working with RMTEC to improve data availability to Tribal nations. We gratefully acknowledge the contributions of Pharah D. Morgan, MS, MPH, and Morgan Witzel, MPH (RMTEC), and we acknowledge Barbara Entl, MD (Office of Science, Medicine and Health, American Health Association), for reviewing the article and offering helpful critiques.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
