NEHA November 2023 Journal of Environmental Health

The November 2023 issue of the Journal of Environmental Health (Volume 86, Number 4), published by the National Environmental Health Association.

JOURNAL OF Environmental Health Dedicated to the advancement of the environmental health professional

Volume 86, No. 4 November 2023

www. neha.org

Published by the National Environmental Health Association

A Comprehensive Solution to Simplify Your Response Process PURELL ® Body Fluid Spill Kits allow for quick and effective protection of employees and guests when an unplanned incident occurs.

Convenience and Compliance • Includes everything needed to safely remove vomit, blood, feces, and other body fluids

• Single-use kits help control cross-contamination • Bilingual instructions with step-by-step images and link to instructional video • Complies with state and local food codes Includes PURELL ® Foodservice Surface Sanitizer • Ready-to-use disinfectant kills norovirus in 30 seconds and the COVID-19 virus in 10 seconds • Formulated for food-contact surfaces with no rinse required • Compatible on a variety of hard and soft surfaces

The Best Defense is a Good Response Plan

GOJO.com/SpillKit

JOURNAL OF Environmental Health Dedicated to the advancement of the environmental health professional Volume 86 No 4 No;ember

ADVANCEMENT OF THE SCIENCE Assessing the Burden of Cold-Related Illness and Death in Minnesota ...................................... 8 Evaluating the Impact of Food and Drug Administration-Funded Cooperative Agreement Programs on Conformance With the Voluntary National Retail Food Regulatory Program Standards .................................................................................................. 16 International Perspectives/Special Report: Bacterial and Viral Pathogens in Drinking Water Sources in Pakistan: A Recent Perspective .......................................................................... 24 ADVANCEMENT OF THE PRACTICE Building Capacity: Building Capacity With the Pragmatic Adoption of Artificial Intelligence ...... 34 Direct From ATSDR: Assessment of Chemical Exposures (ACE) Program: Toolkit Advances and Recent Investigations................................................................................................................. 36 Direct From CDC/Environmental Health Services: Tools From the Centers for Disease Control and Prevention Can Help Prevent and Control Legionella Growth and Spread ................. 40 The Practitioner’s Tool Kit: The Art and Science of Inspection: A Short Introduction .................... 42 Programs Accredited by the National Environmental Health Science and Protection Accreditation Council....................................................................................... 45 ADVANCEMENT OF THE PRACTITIONER Environmental Health Calendar ...............................................................................................46 Resource Corner........................................................................................................................ 47 YOUR ASSOCIATION President’s Message: Environmental Health Data—Can We Make More Powerful Decisions? ................ 6 Special Listing ........................................................................................................................... 48 NEHA 2023 AEC Wrap-Up ....................................................................................................... 50 A Tribute to Our 25-Year and Beyond Members .......................................................................66 U.S. Postal Service Statement of Ownership .............................................................................. 70 NEHA 2024 AEC....................................................................................................................... 71 NEHA News .............................................................................................................................. 72

ABOUT THE COVER

Exposure to cold temperatures can have negative health impacts that lead to cold-related illness or death. This month’s cover

article, “Assess- ing the Burden of Cold-Related

Illness and Death in Minnesota,” explored the case definition for cold-related illness and assessed the burden of cold-related illness and death in Minnesota. As climate change is ex- tending the typical winter season, the authors recommend other jurisdictions consider ex- panding their surveillance window to include all seasons. Cold-related illness surveillance can detect changes over time and identify high- risk populations for prevention initiatives. See page 8. Cover image © iStockphoto: coldsnowstorm

ADVERTISERS INDEX

American Public Health Association (APHA)...... 15 CDP, Inc................................................................ 23 Environmental Health and Land Reuse Certificate Program .............................................. 46 GOJO Industries..................................................... 2 Hedgerow Software ..............................................75 HS GovTech.......................................................... 76 JEH Advertising ....................................................33 NEHA Credentials .................................................. 7 NEHA Endowment Foundation........................... 32 NEHA Membership ................................................ 4 NEHA REHS/RS Credential.................................. 44 NEHA REHS/RS Study Guide............................... 33 NEHA/AAS Scholarship .................................14, 23 NSF......................................................................... 5

November 2023 • our4(l o- 4;0ro4me49(l e(l9/

Join our environmental health community. It is the only community of people who truly understand what it means to do what you do every day to protect the health of our communities. Join us today. Your people are waiting. neha.org/membership Find Your People. Find Your Training. Find Your Resources.

in the next Journal of Environmental Health don’t miss h r(me<or1 -or 88e8804. =6o8ure8 9o ul906le (?(r+8 (4+ m6l0*(90o48 -or r0or090?04. !0818 9o um(4 e(l9/ "(-e9> (4+ 9/e 4;0ro4me49 h +e490->04. &or1-or*e +u*(90o4 #r(0404. (4+ u9re(*/ Nee+8 04 e*e49r(l0?e+ &(89e<(9er (4+ 089r0bu9e+ &(9er !eu8e h roAl04. e9(l4+u*e+ e4o9o=0* Endpoints

Official Publication

Published monthly (except bimonthly in January/February and July/ August) by the National Environmental Health Association, 720 S. Colorado Blvd., Suite 105A, Denver, CO 80246-1910. Phone: (303) 802-2200; Fax: (303) 691-9490; Internet: www.neha.org. E-mail: kruby@neha.org. Volume 86, Number 4. Yearly subscription rate in U.S.: $160 (print). Yearly international subscription rate: $200 (print). Single copies: $15, if available. Reprint and advertising rates available at www.neha.org/jeh. Claims must be filed within 30 days domestic, 90 days foreign, © Copyright 2023, National Environmental Health Association (no refunds). All rights reserved. Contents may be reproduced only with permission of the managing editor. Opinions and conclusions expressed in articles, columns, and other contributions are those of the authors only and do not reflect the policies or views of NEHA. NEHA and the Journal of Environmental Health are not liable or responsible for the accuracy of, or actions taken on the basis of, any information stated herein. NEHA and the Journal of Environmental Health reserve the right to reject any advertising copy. Advertisers and their agencies will assume liability for the content of all advertisements printed and also assume responsibility for any claims arising therefrom against the publisher. The Journal of Environmental Health is indexed by Clarivate, EBSCO (Applied Science & Technology Index), Elsevier (Current Awareness in Biological Sciences), Gale Cengage, and ProQuest. The Journal of Environmental Health is archived by JSTOR (www.jstor.org/journal/ jenviheal). All technical manuscripts submitted for publication are subject to peer review. Contact the managing editor for Instructions for Authors, or visit www.neha.org/jeh. To submit a manuscript, visit http://jeh.msubmit.net. Direct all questions to Kristen Ruby-Cisneros, managing editor, kruby@neha.org. Periodicals postage paid at Denver, Colorado, and additional mailing offices. POSTMASTER: Send address changes to Journal of Environmental Health , 720 S. Colorado Blvd., Suite 105A, Denver, CO 80246-1910.

Journal of Environmental Health (ISSN 0022-0892)

Kristen Ruby-Cisneros, (4(.04. +09or Ellen Kuwana, MS, o6> +09or Hughes design|communications, e80.4 ro+u*90o4 Cognition Studio, o;er r9<or1 Soni Fink, +;er90804. or (+;er90804. *(ll 8 Technical Editors William A. Adler, MPH, RS !e90re+ 044e8o9( e6(r9me49 o- e(l9/ !o*/e89er N Gary Erbeck, MPH

!e90re+ ou49> o- "(4 0e.o e6(r9me49 o- 4;0ro4me49(l e(l9/ "(4 0e.o Thomas H. Hatfield, DrPH, REHS, DAAS (l0-or40( "9(9e $40;er809> Nor9/r0+.e Dhitinut Ratnapradipa, PhD, MCHES re0./9o4 $40;er809> m(/( N

Printed on recycled paper.

Volume 86 • Number 4

On Farm Food Processing Distribution and Retail Food Equipment Dietary Supplements Organic Foods

Life Cycle Analysis Green Building Products Environmental Declarations WaterSense ® Energy Star

Performance and Safety Energy Efficiency Filtration and Recirculation Components

Individual Onsite Wastewater Treatment Systems Advanced Treatment Systems Water Reuse

HACCP Allergens Plan Review SQF, BRC, IFS Food Equipment Traceability and Recall Supply Chain Food Safety

Residential Point-of-Entry/ Point-of-Use Treatment Units Municipal Treatment Chemicals Distribution System Components Plumbing and Devices

Standards • Audits • Testing • Certification Code Compliance • Webinars • Regulatory Support

Visit www.nsf.org/regulatory to submit inquiries, request copies of NSF standards or join the regulatory mailing list.

NSF International • 1-800-NSF-MARK • www.nsf.org/regulatory

November 2023 • Journal of Environmental Health

YOUR ASSOCIATION

 PRESIDENT’S MESSAGE

Environmental Health Data—Can We Make More Powerful Decisions?

Tom Butts, MSc, REHS

W hile we serve our communities, we work in a more and more con- nected environment where shar- ing information and data are required. Doing this sharing in a consistent way, with limited data manipulation, supports better decision making. Environmental public health sys- tems have historically collected a variety of community-, program-, and project-related information and data. These important and potentially useful data have often been placed in spreadsheets, custom databases, or en- terprise software systems designed around workflow, workload management, and ensur- ing regulatory compliance. The information is sometimes publicly available but often kept behind one or more layers of “protection.” There are changes that have occurred with some information that the public and consumers actively sought to access. Retail food inspections are a great example of how demand from the for-profit world has made these data more available and widely used. Initially, big data players (e.g., Yelp and oth- ers) worked to gather these data to add to the information that they provided to their customers and system users who were con- sumers. Now, many (maybe even most) retail food inspections are available on a state or county website, or even shared via social media in near real time. We still have a wide range of data modifiers that are added (e.g., color codes, category descriptors, scoring systems) that often require significant expla- nations and caveats. Community members should be encour- aged to check and understand the narratives or scores of their favorite eateries and patron-

There are many local, state, and national eorts to use program information and data to improve food safety, assure safe practices are adopted, and document regulatory compli- ance. These data are also used on a much more limited basis for academic research, which I suspect is in part due to the wide range of ways the data are collected and the limits around data access. When artificial intelligence (AI) use grows and taps into this information and data, how—for better or worse—will environ- mental health programs, consumers, the pri- vate sector, and even academia be impacted? • Data analysis and decision support: AI algorithms can process large volumes of data quickly and accurately, helping pro- fessionals analyze regulatory requirements, identify patterns, and make informed deci- sions based on the data. It could help with workload analysis and program funding. • Compliance monitoring and risk assess- ment: AI can assist in monitoring and ensuring compliance with regulations by analyzing data from various sources and identifying any anomalies or noncompli- ance activities. It can flag potential issues for further investigation, which can reduce the burden of manual monitoring and increase the eectiveness of regulatory oversight. This process could also assist with workload analysis, fee-for-service justification, or early outbreak risk factor identification that could be addressed with targeted educational outreach. Next, let us consider air quality informa- tion and data. As we work to address air qual- ity impacts from national or international sources, transportation, and point sources

ize those establishments with higher ratings. When data reveal recurring issues in certain establishments, does it prompt targeted inter- ventions or increase consumer interest? I have certainly seen these instances occur. Food safety inspections are not only a formality but also a tool for continuous improvement. Another element of retail food safety data that is of particular interest and importance is the growing reference to one set of stan- dards. It is a program where a national model exists (i.e., the Food and Drug Administra- tion model Food Code ). Data can become more powerful if they are uniform. By con- sistently applying one set of standards, a step toward data standardization is possible. Vari- ous versions of the Food Code from 1995 to 2022 have been adopted in most states (Food and Drug Administration, 2023). These data sources are, however, still fraught with a wide range of implementation models (i.e., vary- ing adoption of the Food Code or state and local variances from the Food Code for local, regional, or governance reasons). As such, there is room for improvement. The eective collection and use of data are crucial for both public health and environmental health initiatives.

Volume 86 • Number 4

in our communities, there is a range of his- torical data and a set of predictive (i.e., lead- ing indicators) that are being used to reduce exposure or impacts. Detailed data on air quality measurements, emissions from indus- tries, and trac patterns can be collected from state and local regulatory agencies and the sources themselves for analysis. Some communities use asthma hospital admissions to document historical impacts. These granular data might allow experts to identify hotspots of pollution and the industries responsible. Provided with this information, communities can work to have these industries adopt cleaner technologies, increase monitoring, and implement warning systems. Inspections of industrial facilities become more stringent and compliance with emission standards can be monitored more closely, which could result in a noticeable drop in air pollution levels. There are many good examples of tools that provide near real-time data available to guide community or individual behaviors. On a national level, the Smoke Forecast- ing System from the National Oceanic and Atmospheric Administration integrates infor- mation on wildfire locations with National Weather Service inputs from the North American Mesoscale model into smoke dis- persion simulations to produce a daily 48-hr prediction of smoke transport and concentra- tion. The model also incorporates U.S. Forest Service estimates for wildfire smoke emis- sions based on vegetation cover. This system is intended as guidance to air quality forecast-

ers and the public for fine particulate matter emitted from large wildfires and agricultural burning that can elevate particulate concen- trations to unhealthful levels. The system is a great near real-time resource for decision making within environmental public health (https://digital.mdl.nws.noaa.gov/airquality). On a local level, data on community water quality for cyanobacteria (also known as blue-green algae) in Vermont is collected by regulatory agencies and citizen scientists on an ongoing basis. These online reports are continually updated and are then displayed on the Cyanobacteria (Blue-Green Algae) Tracker (www.healthvermont.gov/environ ment/tracking/cyanobacteria-blue-green- algae-tracker). This resource can be used by individuals as well as water resource manag- ers and health ocials. This local example is just one of many data sources available via the National Envi- ronmental Public Health Tracking Network (www.cdc.gov/nceh/tracking/index.html). At local, state, and national levels, the Tracking Network uses groups of people and informa- tion systems to deliver a core set of health, exposure, and hazards data; information summaries; and tools to enable analysis, visu- alization, and reporting of insights drawn from data. As discussed above, gathering the data from a wide range of sources and sys- tems, and getting it into a usable form, is a large part of the eort to make these data available and useful. The eective collection and use of data are crucial for both public health and envi-

ronmental health initiatives. Environmental public health can benefit immensely from data-driven decision making. By implement- ing these practical strategies, local commu- nity needs, and national initiatives, we can better understand the unique challenges and work toward creating healthier and more sus- tainable environments. Data alone will not be enough to inform community members or elected ocials about these challenges and needs. Relatable stories must accompany the data to create and support the case for change or program improvement. On a final note, the Building Capacity column in the September 2023 Journal of Environmental Health provided a nicely writ- ten and thought-provoking discussion about generative AI considerations (www.neha.org/ Images/resources/JEH9.23-Column-Build ing-Capacity.pdf). You can also find a new Building Capacity column in this issue that explores programmatic AI adoption.

tbutts@neha.org

Reference Food and Drug Administration. (2023). Adoption of the FDA Food Code by state and territorial agencies responsible for the over- sight of restaurants and retail food stores . https://www.fda.gov/food/fda-food-code/ adoption-fda-food-code-state-and-territor ial-agencies-responsible-oversight-restau rants-and-retail

Stand out in the crowd. Show the world you are the environmental health expert you know you are with a credential. You might even earn more or get promoted. neha.org/credentials

November 2023 • our4(l o- 4;0ro4me49(l e(l9/

VNN# # SCIENCE

Assessing the Burden of Cold-Related Illness and Death in Minnesota

Madison Kircher, MPH Tess Konen, MPH Jessie Carr, MPH, DrPH

Environmental Health Division, Minnesota Department of Health

illness and death that capture a broad range of known risk factors and exposure circum- stances, as well as emergent climate change- related conditions, using methods that can be duplicated in other jurisdictions. Cold-related illness occurs when the body loses heat faster than it can be produced. This category of conditions includes hypothermia (a reduction in the body’s core temperature to below 95 °F [35 °C]) and injuries such as frostbite, trench foot, or chilblains (skin sores or bumps that occur after exposure to cold temperatures but rarely cause perma- nent damage). While these conditions are most likely to occur due to prolonged expo- sure to subfreezing temperatures (i.e., <32 °F), they can occur at temperatures as high as 40 °F in wind or rain, or 70 °F in some individuals with underlying medical condi- tions (Nixdorf-Miller et al., 2006). Most previous research has focused on the impacts of climate change on heat-related illness, with few studies describing vulner- able populations or contributing factors to cold-related illness and death. Similar to heat-related illness, infants, older adults (>65 years), and individuals with specific chronic conditions (e.g., respiratory disease, cardio- vascular disease) are more susceptible to cold-related illness and death (Berko et al., 2014; Gronlund et al., 2018; Nixdorf-Miller et al., 2006). Individuals who consume alco- hol, take illicit drugs, or use some medica- tions are also more susceptible, as these sub- stances can adversely aect the body’s ability to sense the cold (Gronlund et al., 2018; Nixdorf-Miller et al., 2006). One study found that hyperthermia- related visits were more frequent than hypothermia-related visits among Medicare

b89r(*9 Exposure to cold temperatures can have negative health impacts that lead to cold-related illness or death. We explored the case definition for cold-related illness that was developed and piloted by the National Environmental Public Health Tracking Network within the Centers for Disease Control and Prevention. Using their case definition, we assessed the burden of cold-related illness and death in Minnesota. We analyzed the results by season, demographics, and chronic disease. Overall, <10% of all cold-related events in Minnesota occurred during the hot season; we did not identify any distinct dierences between the type of cases by seasons. During the cold season, there was an average annual rate of 13.3 cold-related emer- gency department visits per 100,000 population ( n = 704) and 2.8 cold-relat- ed hospitalizations per 100,000 population ( n = 155). There was an average annual rate of 0.6 cold-related deaths per 100,000 population ( n = 33). Cli- mate change is extending the typical winter season. Therefore, we recommend other jurisdictions consider expanding their surveillance window to include all seasons. Cold-related illness surveillance can detect changes over time and identify high-risk populations for prevention initiatives.

Introduction Cold-related illness and death are common and occur across dierent U.S. regions. The National Center for Health Statistics within the Centers for Disease Control and Preven- tion (CDC) found that almost two thirds (63%) of all mortality coded as weather- related from 2006 to 2010 in the U.S. was due to cold exposure, while less than one third (31%) was attributable to heat expo- sure (Berko et al., 2014). While climate change is contributing to increasing average winter temperatures, cold-related illness and death will continue to be health risks. One study found that most cold-related mortality

was caused by exposure to moderately cold temperatures, but that the contribution of extremely cold temperatures was compara- tively low, suggesting that reductions in cold- related mortality from climate change might be smaller than initially assumed (Gasparrini et al., 2015). Climate change can have other impacts on cold weather, such as increases in the inten- sity of extreme cold events and winter storms, which in turn have important implications for cold-related illness and death (Conlon et al., 2011; Noe et al., 2012). The purpose of our assessment was to develop and evaluate locally relevant surveillance measures for cold-related

Volume 86 • Number 4

claims data (Noe et al., 2012). Hypothermia resulted in higher mortality rates, longer hos- pital stays, and higher total healthcare costs, however, indicating an increased burden of cold-related illness and death among older adults (Noe et al., 2012). Other studies have found higher rates of cold-related illness and death among men and individuals experienc- ing homelessness. A study from New York City (Lane et al., 2018) found that men, older adults, and those with multiple chronic conditions were more likely to be hospitalized or die due to cold exposure compared with those treated and released from the emergency depart- ment (ED). The most common chronic conditions found among those hospitalized with cold-related illness included cardiovas- cular disease, substance use, and mental ill- ness (Lane et al., 2018). The majority of the state of Minnesota is located in the humid continental climate zone, a zone that is characterized by hot summers and cold winters (Peel et al., 2007). Despite being in a cold weather climate zone, however, cold-related illness and death have not been systematically monitored by the Minnesota Department of Health. The Cold- Related Illness Content Work Group within the CDC Environmental Public Health Track- ing Network piloted the case definition in Kentucky, Massachusetts, New Jersey, New Mexico, New York, Vermont, and Wisconsin, as well as in New York City. The case defi- nition excluded events occurring during the hot season—defined as the months of May through September—due to evidence from the pilot testing suggesting that events in the hot season were related to cold water expo- sure rather than cold temperature. Our assessment evaluated the utility of this case definition for cold-related illness and death in Minnesota. The findings from this assessment can be used to identify vulner- able populations and develop targeted inter- ventions to prevent adverse outcomes from cold exposure in Minnesota and inform other jurisdictions about monitoring cold-related illness and death. Methods Cold-related illness and death in Minnesota were assessed using the International Classi- fication of Diseases (ICD), 9th and 10th Revi- sions, Clinical Modification (ICD-9-CM and

ICD-10-CM) codes from the case definition developed by the Cold-Related Illness Content Work Group. We examined Minnesota Hospi- tal Discharge Data (MNHDD) for ED visits and hospitalizations for cold-related illness from 2000 to 2018, the period for which complete data were available. MNHDD is a comprehen- sive data set that includes patient-level claims data from the majority of hospital visits in the state (excluding the Minnesota Department of Veterans Aairs and Indian Health Service). We defined ED and hospital cases as patients with any ICD-9-CM diagnosis code of 991 (“eects of reduced temperature”); external cause of injury code E901.0, E901.8, E901.9, or E988.3 (“excessive cold” or “extremes of cold” of unintentional or unde- termined intent); or ICD-10-CM code of X31, T68, T69, T33, or T34 (“exposure to exces- sive natural cold,” “hypothermia,” “other eects of reduced temperature,” “superficial frostbite,” or “frostbite with tissue necrosis”) in any diagnosis field. We excluded records with any diagnosis of ICD-9-CM E901.1 or ICD-10-CM W93 (“excessive cold of human- made origin”) and non-Minnesota residents. Cold-related deaths occurring from 2002 to 2019, the period for which complete data were available, were examined using death certificate data provided by the Minnesota Center for Health Statistics at the Minnesota Department of Health. Cases were defined as deaths among Minnesota residents with an ICD-10-CM code of X31, T68, T69, T33, or T34 as an underlying or contributing cause of death. We excluded records with any diag- nosis of ICD-10-CM W93 and intentional deaths. We also excluded any out-of-state deaths, as we included only Minnesota death certificate records from Minnesota residents in our analysis. The case definition for cold-related illness was developed and piloted by the Council of State and Territorial Epidemiologists and the Cold-Related Illness Content Work Group. We explored this case definition by examin- ing the proportion and type of events that occurred outside of the cold season. The cold season was defined as January–April and October–December, and the hot sea- son was defined as May–September. We also explored the hypothesis that cases in the hot season might be related to cold water expo- sure. Water-related ICD-10 codes included W69, W70, and W74 (“accidental drown-

ing and submersion while in natural water,” “drowning and submersion following fall into natural water,” and “unspecified cause of accidental drowning and submersion”). After examining the proportion and type of cases in the summer months, we calculated rates for the cold months using the current winter season case definition. We conducted descriptive statistics for cold-related illness and death in Minnesota. The annual number and rate of cold-related ED visits, hospitalizations, and deaths were calculated by age and sex. Race data were incomplete and homogeneously White. The most recent 5 years of data were aggregated for cold-related ED visits and hospitaliza- tions, while 10-year aggregated data were used for cold-related deaths. We extracted Minnesota population esti- mates for the relevant years from the U.S. Census Bureau and American Community Survey. Age-adjusted rates were calculated using the direct method and the U.S. 2000 standard population. We compared rates of cold-related illness and death across sex and age groups using variance testing (ANOVA) with post hoc Tukey tests. Statistical signifi- cance was defined as p < .05. We also exam- ined the prevalence of cardiovascular dis- ease, respiratory conditions, substance use, mental illness, and diabetes that co-occurred with the cold-related diagnosis, as these conditions are known contributing factors for cold-related illness and death (Berko et al., 2014; Gronlund et al., 2018; Lane et al., 2018; Nixdorf-Miller et al., 2006).

Results

Surveillance Window Approximately 1 in 10 (10%) cold-related ED visits and hospitalizations from 2000 to 2018 occurred during the hot season, while >90% of cold-related ED visits and hospitalizations occurred during the cold season (Table 1). Similarly, only 6% of cold-related deaths from 2002 to 2019 occurred during the hot season, compared with 94% during the cold season. During the hot season, the highest propor- tion of ED visits, hospitalizations, and deaths occurred in May. Overall, the highest propor- tion of cold-related illness and death in any month occurred during January. The type of cold-related illness and death events in the hot season were similar to

November 2023 • our4(l o- 4;0ro4me49(l e(l9/

ADVANCEMENT OF THE SCIENCE

events in the cold season. We found that <1% of ED visits were water-related during the cold season and no water-related ED vis- its or hospitalizations were identified dur- ing the hot season. Additionally, only 2% of deaths were water-related in the cold season, while 6% of deaths in the hot season were water-related. There were no other clear dis- tinctions between the type of events occur- ring during the hot and cold seasons. For the remaining analysis, we used the case definition implemented by the CDC Envi- ronmental Public Health Tracking Network, which restricts the definition to include only cold-season cases. Hospital Visits During each cold season from 2000 to 2018, there was an average rate of 13.3 cold-related ED visits per 100,000 population ( n = 704) and 2.8 cold-related hospitalizations per 100,000 population ( n = 155). The annual rate of cold-related ED visits and hospital- izations has been trending upward in recent years (Figures 1 and 2). The highest rate of cold-related hospitalizations during this time period occurred in 2018 (Figure 2). Overall, there were more cold-related ED visits than hospitalizations for the years analyzed. Females accounted for approximately 30% of cold-related ED visits and hospitalizations, while males accounted for 70% (Table 2). There was a statistically significant dier- ence between the sex distribution of the rate of cold-related ED visits and hospitalizations. For age distributions by sex, males 15–34 years had the highest rates of cold-related ED visits, while males ≥65 years had the highest rates of cold-related hospitalizations (Table 2). Among females, there was a sta- tistically significant dierence between the rate of cold-related ED visits for the 15–34- year group and all other age groups. Among males, the rate of cold-related ED visits was significantly higher for the 15–34-year group compared with the 0–4, 5–14, and ≥65 age groups. For cold-related hospital- izations, there was a statistically significant dierence between the rates for the ≥65- year group compared with the other age groups for females. For males, the 15–34, 35–64, and ≥65 age groups had significantly higher hospitalization rates compared with the other age groups. There was no statisti- cally significant dierence between the rates

TABLE 1

Number and Proportion of Cold-Related Events by Month in Minnesota

Month

Season

Emergency Department Visits, 2000–2018 # (%)

Hospital Admissions, 2000–2018 # (%)

Deaths, 2002–2019 # (%)

January February

Cold

4,055 (28) 2,809 (19)

844 (26) 651 (20) 354 (11)

160 (26) 93 (15) 91 (15)

March

1,372 (9)

April May June

603 (4) 380 (3) 250 (2) 185 (1) 197 (1) 273 (2) 577 (4)

158 (5)

33 (5) 19 (3)

Hot

99 (3) 60 (2) 43 (1) 49 (1) 60 (2)

6 (1)

July

3 (<1) 3 (<1)

August

September

5 (1)

October

Cold

132 (4) 253 (8) 546 (17)

38 (6)

November December

1,043 (7)

65 (10)

2,917 (20)

108 (17)

Total

14,661 (100)

3,249 (100)

624 (100)

of cold-related hospitalizations or ED visits in the age groups of 0–4 or 5–14 years for males or females. Almost one half of the cold-related ED visits (45%) included diagnosis codes for substance use (Table 3). Other diagnosis codes co-occur- ring with cold-related ED visits included men- tal illness (11%), respiratory disease (8%), car- diovascular disease (7%), and diabetes (7%). Almost all the cold-related hospitalizations had at least one co-occurring diagnosis code (89%), including substance use (66%), men- tal illness (33%), respiratory disease (22%), or cardiovascular disease (20%; Table 3). Deaths We identified an average annual rate of 0.6 cold-related deaths per 100,000 population ( n = 33) over each cold season from 2002 to 2019. Similar to the hospital discharge data, there was a statistically significant dif- ference between the sex distribution of cold- related deaths, with females accounting for approximately 30% of cold-related deaths, while males accounted for 70% (Table 2). For both males and females, there was a statisti- cally significant dierence between the rate of cold-related deaths for the ≥65 age group compared with all other age groups.

More than one half of all cold-related deaths (57%) had co-occurring diagnosis codes (Table 3). Almost one half of all cold- related deaths (44%) included a diagnosis code for substance use. Other co-occurring diagnosis codes included cardiovascular dis- ease (19%), respiratory disease (8%), mental illness (2%), and diabetes (1%). Discussion Our study used hospital discharge data and vital statistics data to explore the case defi- nition and assess the burden of cold-related illness and death in Minnesota. Overall, <10% of cold-related ED visits, hospitaliza- tions, and deaths in Minnesota occurred during the hot season. We were unable to identify any distinct dierence between the type of events occurring in the dierent seasons. We assessed the burden of cold- related illness and death in Minnesota using the case definition developed and adopted by the Cold-Related Illness Content Work Group, which includes cases only in the cold season. Using this case definition, we found that rates of illness and death in Min- nesota were highest among older adults and males, which is consistent with previ- ous studies (Gronlund et al., 2018; Lane

Volume 86 • Number 4

et al., 2018; Nixdorf-Miller et al., 2006). Additionally, the most common co-occur- ring diagnosis with cold-related illness and death was substance use. We hypothesized that cases in the hot season might be related to cold water expo- sure rather than cold weather or air tem- perature exposure. There were very few cases with water-related ICD codes, however, that occurred in the cold or hot seasons in Minne- sota. Hypothermia could be due to cold water exposure in addition to cold temperature exposure; it is also possible that water-related hypothermia cases did not get properly docu- mented with the ICD codes to indicate that water exposure was involved. Our analysis also found that the highest proportion of cases in the hot season occurred during May, suggesting that a possible next step could involve expanding the definition to include this “shoulder-season” month. As climate change continues to disrupt patterns and distribution of rain and snow, we could see more snowfall outside of the typical cold season, further emphasizing the importance of expanding the surveillance window to include events in the hot season. Based on these find- ings, we recommend that other jurisdictions explore and present data on cold-related ill- ness and death using both the case definition restricted to the cold season and the case defi- nition that includes cases year-round. In Minnesota, there were almost 2 times more cold-related illness ED cases than heat- related illness during the most recent 5 years of data (Minnesota Department of Health, n.d.). Both conditions had the same high- risk group profile of ED visits: highest among males 15–34 years and hospitalizations highest among males ≥65 years (Minnesota Department of Health, n.d.). Additionally, there were more cold-related deaths annually compared with heat-related deaths during the study period, which is consistent with existing research comparing hyperthermia and hypothermia (Noe et al., 2012). Previous studies have also found that hyperthermia deaths were related to extreme heat events, while most cold-related deaths occurred on days that were colder than aver- age, but not extremely cold—suggesting that it is important to prevent exposure to the cold even when the temperatures are not extreme (Gasparrini et al., 2015; Gronlund et al., 2018). Additional research could assess this

FIGURE 1

Number and Rate of Cold-Related Illness Emergency Department Visits in Minnesota by Year, 2000–2018

1 ,60 0

3 0

1 ,40 0

2 5

1 ,2 0 0

2 0

1 ,0 0 0

80 0

1 5

60 0

1 0

40 0

2 0 0

2 0 0 0 2 0 0 1 2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 2 0 0 6 2 0 0 7 2 0 0 8 2 0 0 9 2 0 1 0 2 0 1 1 2 0 1 2 2 0 1 3 2 0 1 4 2 0 1 5 2 0 1 6 2 0 1 7 2 0 1 8 # of Emergency Department V isits Age-Adjusted Rate per 1 0 0 ,0 0 0

Note. Rates from 2000–2014 should not be compared with rates from 2015 onward due to a change in the International Classification of Diseases (ICD) coding from ICD-9 to ICD-10 on October 1, 2015. Source: Minnesota Environmental Public Health Tracking Program data access portal (https://data.web.health.state. mn.us/web/mndata/cold_related_illness).

FIGURE 2

Number and Rate of Cold-Related Illness Hospitalizations in Minnesota by Year, 2000–2018

3 50

3 0 0

2 50

2 0 0

0 1 2 3

1 50

1 0 0

2 0 0 0 2 0 0 1 2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 2 0 0 6 2 0 0 7 2 0 0 8 2 0 0 9 2 0 1 0 2 0 1 1 2 0 1 2 2 0 1 3 2 0 1 4 2 0 1 5 2 0 1 6 2 0 1 7 2 0 1 8 # of Hospitalizations Age-Adjusted Rate per 1 0 0 ,0 0 0

November 2023 • Journal of Environmental Health

ADVANCEMENT OF THE SCIENCE

TABLE 2

Number, Proportion, and Rate of Cold-Related Events by Sex and Age Group in Minnesota

Emergency Department Visits, 2014–2018

Hospital Admissions, 2014–2018

Deaths, 2010–2019

# (%)

Rate per 100,000

95% CI

# (%)

Rate per 100,000

95% CI

# (%)

Rate per 100,000

95% CI

Total

5,227 (100)

–

1,048 (100)

–

426 (100)

–

Sex

Female

1,546 (30) 3,681 (70)

11.4

[10.8, 11.9] [26.3, 28.0]

300 (29) 748 (71)

2.0

[1.8, 2.2] [5.3, 6.1]

134 (31) 292 (69)

0.4

[0.3, 0.5] [0.9, 1.2]

Male

27.2 a

5.7 a

1.0 a

Female age group (years) 0–4

50 (3)

5.8 6.3

[4.3, 7.7] [5.1, 7.5]

9 (3) 3 (1)

1.0 c 0.2 c

[0.5, 2.0]

0 (0) 2 (0)

5–14

112 (7)

[0, 0.5]

0.1 c 0.2 c

[0, 0.2]

15–34 35–64

650 (42) 502 (33) 232 (15)

18.0 a

[16.6, 19.4] [8.6, 10.2] [8.8, 11.5]

69 (23)

1.9 2.3

[1.5, 2.4] [1.9, 2.7] [3.4, 5.2] [0.3, 1.6] [0.1, 0.6] [4.4, 5.9] [7.2, 8.7] [6.8, 9.4]

14 (10) 44 (33) 74 (55)

[0.1, 0.3] [0.3, 0.6] [1.3, 2.1]

9.4

122 (41)

0.4

≥65

10.2

97 (32)

4.2 a

1.7 a

Male age group (years) 0–4

62 (2)

6.9 6.8

[5.3, 8.8] [5.6, 7.9]

7 (1) 4 (1)

0.8 c 0.2 c 5.2 b 7.9 b 8.1 b

0 (0) 1 (0)

5–14

125 (3)

0 c

[0, 0.2]

15–34 35–64

1,422 (39) 1,702 (46)

37.9 b 31.6 b

[35.9, 39.9] [30.1, 33.1] [17.7, 21.8]

194 (25) 427 (55) 152 (19)

40 (14)

0.5 1.3

[0.4, 0.7] [1.1, 1.5] [2.6, 3.8]

137 (47) 114 (39)

≥65

370 (10)

19.8

3.2 a

Note. Data are restricted to cold-related events occurring in January to April and October to December. Rates are calculated using 2010 U.S. Census Bureau data for the denominators. CI = confidence interval. a Significantly higher than other groups ( p < .05). b Significantly higher than other groups but not significantly different from each other ( p < .05). c Rates based on counts <20 are flagged as unstable because they can change dramatically with the addition or subtraction of one case.

relationship further by exploring other con- tributing factors involved in cold-related ill- ness and death outside of temperature, such as occupational and social risk factors. Chronic conditions, such as substance use, co-occurred with cold-related illness and death in Minnesota, which is consistent with previous studies (Berko et al., 2014; Gronlund et al., 2018; Lane et al., 2018). Substance use can adversely aect the body’s ability to sense the cold and can cloud decision making, par- tially explaining this relationship. In Minne- sota, rates of drug overdoses and deaths have been increasing, which might be contributing to the rise in cold-related illness and death in recent years (DeLaquil et al., 2020). Substance use is both a cause and conse- quence of homelessness. Homelessness is an additional risk factor for cold-related ill- ness and death, which has been on the rise in Minnesota (Minnesota Department of

Health, 2023). People with mental health conditions might also be at increased risk for cold-related illness and death in part due to psychiatric medications that can impair ther- moregulation (Gronlund et al., 2018). While mental illness was listed on only 2% of death certificate records, a similar study from New York City found that it was noted on a higher proportion of decedents in medical examiner records (Lane et al., 2018). Future studies in Minnesota could explore medical examiner records to obtain more detailed information on contributing factors in cold-related deaths. There are several limitations to our analy- sis. These data only captured the individuals with the most severe or acute symptoms who were treated at the ED, hospitalized, or die. Thus, the cases likely are underestimated. We have provided a descriptive analysis of condi- tions co-occurring with a cold-related diag- nosis and recommend that a more rigorous

analysis be completed to elucidate the rela- tionship between these conditions and cold- related illness. Additionally, we do not have access to the full health records or patient history for individual hospitalizations and deaths, so we could be missing important contextual information. The data on hospitalization and death lack information on social factors, unhoused status, and occupation-related exposure, all of which would provide more insight into understanding the risk factors and context for the cases. The Minnesota Department of Health is expanding its syndromic surveillance pro- gram and plans to pilot the cold-related ill- ness case definition. These data often include more details and context that can give us a better understanding of the risk factors, high- risk groups, and geographical patterns of cold-related illness.

Volume 86 • Number 4

and death outside of the standard definition that is limited to the cold season.

TABLE 3

Acknowledgements: The authors acknowledge the Minnesota Environmental Public Health Tracking Program and Minnesota Climate & Health Program for their assistance with this project. We also acknowledge the support and guidance of the Council of State and Territorial Epidemiologists and the CDC Environmental Public Health Tracking Network Cold-Related Illness Content Work Group. This work was supported in part by an appointment to the Applied Epidemiology Fellowship Program administered by the Council of State and Territorial Epidemiolo- gists and funded by CDC Cooperative Agree- ment Number 1NU38OT000297-03-00. This work was also supported by the Minnesota Environmental Public Health Tracking Pro- gram, grant number CDC-RFA-EH17-1702, funded by CDC. The findings and conclu- sions are solely those of the authors and do not necessarily reflect the views of CDC or the Minnesota Department of Health. Corresponding Author: Tess Konen, Senior Epidemiologist, Environmental Health Divi- sion, Minnesota Department of Health, 625 Robert Street North, St. Paul, MN 55155. Email: tess.konen@state.mn.us.

Number and Proportion of Other Health Conditions That Co-occur With Cold-Related Illness or Death in Minnesota

Other Health Conditions

Emergency Department Visits, 2015–2018 # (%)

Hospitalizations, 2015–2018 # (%)

Deaths, 2010–2019 # (%)

Any chronic condition Cardiovascular disease

1,878 (56)

664 (89) 151 (20) 100 (14) 247 (33) 165 (22) 492 (66)

241 (57)

236 (7) 234 (7) 382 (11)

80 (19)

Diabetes

6 (1) 9 (2)

Mental illness

Respiratory disease

273 (8)

35 (8)

Substance use

1,487 (45)

187 (44)

Note. Emergency department visits and hospitalizations are not mutually exclusive. Any chronic condition is defined as having one or more of the following conditions: cardiovascular disease, substance use, mental illness, respiratory disease, or diabetes.

Conclusion Our analysis examined the case definition and assessed the burden of cold-related illness and death in Minnesota. Despite warming winter temperatures due to climate change, cold- related illness and death will continue to be health risks. Additional research and discus- sion are needed to inform decision making about expanding the surveillance window, but

we recommend that jurisdictions explore the case definition in both ways: restricted to the winter season and year-round cases. By assess- ing cold-related illness and death, changes in the distribution can be detected, high-risk groups can be monitored, and prevention ini- tiatives can be developed. This study provides a locally relevant analytic framework for other jurisdictions to evaluate cold-related illness

References

Berko, J., Ingram, D.D., Saha, S., & Parker, J.D. (2014). Deaths attrib- uted to heat, cold, and other weather events in the United States, 2006–2010. National Health Statistics Reports , 76 , 1–15. Conlon, K.C., Rajkovich, N.B., White-Newsome, J.L., Larsen, L., & O’Neill, M.S. (2011). Preventing cold-related morbidity and mor- tality in a changing climate. Maturitas , 69 (3), 197–202. https:// doi.org/10.1016/j.maturitas.2011.04.004 DeLaquil, M., Giesel, S., & Wright, N. (2020). Drug overdose deaths among Minnesota residents, 2000–2018 . Minnesota Department of Health. https://www.lrl.mn.gov/docs/2021/other/210532.pdf Gasparrini, A., Guo, Y., Hashizume, M., Lavigne, E., Zanobetti, A., Schwartz, J., Tobias, A., Tong, S., Rocklöv, J., Forsberg, B., Leone, M., De Sario, M., Bell, M.L., Guo, Y.-L.L, Wu, C.-F., Kan, H., Yi, S.-M., de Sousa Zanotti Stagliorio Coelho, M., Saldiva, P.H.N., . . . Armstrong, B. (2015). Mortality risk attributable to high and low ambient temperature: A multicountry observa- tional study. Lancet , 386 (9991), 369–375. https://doi.org/10.1016/ s0140-6736(14)62114-0

Gronlund, C.J., Sullivan, K.P., Kefelegn, Y., Cameron, L., & O’Neill, M.S. (2018). Climate change and temperature extremes: A review of heat- and cold-related morbidity and mortality concerns of municipalities. Maturitas , 114 , 54–59. https://doi.org/10.1016/j. maturitas.2018.06.002 Lane, K., Ito, K., Johnson, S., Gibson, E.A., Tang, A., & Matte, T. (2018). Burden and risk factors for cold-related illness and death in New York City. International Journal of Environmental Research and Public Health , 15 (4), Article 632. https://doi.org/10.3390/ ijerph15040632 Minnesota Department of Health. (n.d.). Heat-related illness: How weather can be deadly . https://data.web.health.state.mn.us/web/ mndata/heat Minnesota Department of Health. (2023). Minnesota homeless mor- tality report, 2017–2021 . https://www.health.state.mn.us/commu nities/homeless/coe/coephhmr.pdf

continued on page 14

November 2023 • our4(l o- 4;0ro4me49(l e(l9/

Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15 Page 16 Page 17 Page 18 Page 19 Page 20 Page 21 Page 22 Page 23 Page 24 Page 25 Page 26 Page 27 Page 28 Page 29 Page 30 Page 31 Page 32 Page 33 Page 34 Page 35 Page 36 Page 37 Page 38 Page 39 Page 40 Page 41 Page 42 Page 43 Page 44 Page 45 Page 46 Page 47 Page 48 Page 49 Page 50 Page 51 Page 52 Page 53 Page 54 Page 55 Page 56 Page 57 Page 58 Page 59 Page 60 Page 61 Page 62 Page 63 Page 64 Page 65 Page 66 Page 67 Page 68 Page 69 Page 70 Page 71 Page 72 Page 73 Page 74 Page 75 Page 76