Public Health at BMJ
Table of contents Epidemiology of SARS-CoV-2 antibodies among firefighters/paramedics of a US fire department: a cross-sectional study (Occupational & Environmental Medicine) Vulnerability of the medical product supply chain: the wake-up call of COVID- 19 (BMJ Quality & Safety) Changes in intimate partner violence during the early stages of the COVID-19 pandemic in the USA (Injury Prevention) Racial inequity in fatal US police shootings, 2015–2020 (Journal of Epidemiology and Community Health)
Workplace
SHORT REPORT Epidemiology of SARS-CoV-2 antibodies among firefighters/paramedics of a US fire department: a cross-sectional study Alberto J Caban-Martinez , 1,2 Natasha Schaefer-Solle, 2,3 Katerina Santiago, 1 Paola Louzado-Feliciano, 1 Angel Brotons, 4 Marco Gonzalez, 4 S. Barry Issenberg, 4 Erin Kobetz 1,2,3
1 Public Health Sciences, University of Miami Miller School of Medicine, Miami, Florida, USA 2 Sylvester Comprehensive Cancer Center, Miami, Florida, USA 3 Medicine, University of Miami School of Medicine, Miami, Florida, USA 4 Gordon Center for Research in Medical Education, University of Miami Miller School of Medicine, Miami, Florida, USA Correspondence to Dr Alberto J Caban-Martinez, Public Health Sciences, University of Miami Miller School of Medicine, Miami, FL 33136, USA; a caban@med.m iami.edu
ABSTRACT Objectives We estimate the point seroprevalence of SARS-CoV-2 antibodies in the frontline firefighter/ paramedic workforce of a South Florida fire department located in the epicentre of a State outbreak. Methods A cross-sectional study design was used to estimate the point seroprevalence of SARS-CoV-2 antibodies using a rapid immunoglobulin (Ig)M-IgG combined point-of-care lateral flow immunoassay among frontline firefighters/paramedics collected over a 2-day period, 16–17 April 2020. Fire department personnel were emailed a survey link assessing COVID-19 symptoms and work exposures the day prior to the scheduled drive-through antibody testing at a designated fire station. Off-duty and on-duty firefighter/paramedic personnel drove through the fire station/training facility in their personal vehicles or on-duty engine/rescue trucks for SARS-CoV-2 antibody testing. Results Among the 203 firefighters/paramedics that make up the fire department workforce, 18 firefighters/ paramedics (8.9%) tested positive for SARS-CoV-2 antibodies, of which 8 firefighters/paramedics (3.9%) were IgG positive only, 8 (3.9%) were IgM positive only and 2 (0.1%) were IgG/IgM positive. The positive predictive value (PPV) of the serological test is estimated to be 33.2% and the negative predictive value is 99.3%. The average number of COVID-19 case contacts (ie, within 6 feet of an infected person (laboratory- confirmed or probable COVID-19 patient) for ≥15 min) experienced by firefighters/paramedics was higher for those with positive serology compared with those with negative (13.3 cases vs 7.31 cases; p=0.022). None of the antibody positive firefighters/paramedics reported receipt of the annual influenza vaccine compared with firefighters/paramedics who tested negative for SARS- CoV-2 antibodies (0.0% vs 21.0%; p=0.027). Conclusion Rapid SARS-CoV-2 IgM-IgG antibody testing documented early-stage and late-stage infection in a firefighter workforce providing insight to a broader medical surveillance project on return to work for firefighters/paramedics. Given the relatively low PPV of the serological test used in this study back in April 2020, caution should be used in interpreting test results.
Key messages
first responders from coronavirus evolved following the first American COVID-19 case and the exposure of at least one firefighter. 1 Among all US jobs, those employed as first responders, that is, firefighters/ paramedics, are at greatest risk for COVID-19 infection, as they can encounter diseases and infec- tions daily and typically work in close proximity to one another and the communities they serve. 2 Many first responders are already under quarantine due to direct exposure with COVID-19 cases, potentially challenging fire department staffing resources and emergency responder workforce responsiveness. 3 While firefighters/paramedics use personal protec- tive equipment (PPE) and engineering controls at work, fire departments are operating in the dark ► ► None of the antibody positive firefighters/ paramedics reported receipt of the annual influenza vaccine compared with firefighters/ paramedics who tested negative for SARS- CoV-2 antibodies (0.0% vs 21.0%; p=0.027). How might this impact on policy or clinical practice in the foreseeable future? ► ► A comprehensive medical surveillance programme for first responders that includes SARS-CoV-2 antibody testing can inform policy for return to work algorithms. What is already known about this subject? ► ► Among all occupations, those employed as first responders, that is, firefighters/paramedics, are at greatest risk for COVID-19 infection, as they can encounter diseases and infections daily and typically work in close proximity to one another and the communities they serve. What are the new findings? ► ► We found the seroprevalence of SARS-CoV-2 antibodies (immunoglobulin (Ig)G only, IgM only or IgG/IgM) estimated in a cross-sectional study of 203 frontline firefighters/paramedics from a municipal fire department was 8.9% of the workforce, of which eight firefighters/ paramedics (3.9%) were IgG positive only, eight (3.9%) were IgM positive only and two (0.1%) were IgG/IgM positive.
Received 7 May 2020 Revised 12 July 2020 Accepted 20 July 2020 Published Online First 6 August 2020
© Author(s) (or their employer(s)) 2020. No
To cite: Caban-Martinez AJ, Schaefer-Solle N, Santiago K, et al . Occup Environ Med 2020; 77 :857–861. commercial re-use. See rights and permissions. Published by BMJ.
INTRODUCTION Key direction from the US Centers for Disease Control and Prevention (CDC) on how to protect
Caban-Martinez AJ, et al . Occup Environ Med 2020; 77 :857–861. doi:10.1136/oemed-2020-106676
857
Workplace
Table 1 Sociodemographic and work characteristics among firefighters who participated in voluntary serological COVID-19 antibody test (n=203)
COVID-19 antibody test result
Total sample N N (%)*
Positive (IgG, IgM or IgG/IgM) n (%)*
Negative n (%)*
Characteristics
P value
Total
203 (100.0)
18 (8.9)
185 (91.1)
Age groups
0.677
21–30 years old 31–40 years old 41–50 years old 51 years and older
33 (16.3) 51 (25.1) 67 (33.0) 52 (25.6)
2 (11.1) 6 (33.3) 7 (38.9) 3 (16.7)
31 (16.8) 45 (24.3) 60 (32.4) 49 (26.5)
Sex
0.328
Male
188 (93.5)
16 (88.9) 2 (11.1)
172 (94.0)
Female
13 (6.5)
11 (6.0)
Race
0.899
White
154 (78.2)
15 (83.3) 0 (0.0) 1 (5.6) 2 (11.1)
139 (77.7)
Black or African–American
9 (4.6) 8 (4.1)
9 (5.0) 7 (3.9)
Multi-race
Other
26 (13.2)
24 (13.4)
Ethnicity
0.57
Hispanic/Latinx
149 (75.6) 48 (24.4)
15 (83.3) 3 (16.7)
134 (74.9) 45 (25.1)
Non-Hispanic/non-Latinx
Marital status
0.721
Married/unmarried couple Divorced, widowed, separated
139 (72.4) 21 (10.9) 32 (16.7)
12 (66.7) 2 (11.1) 4 (22.2)
127 (73.0) 19 (10.9) 28 (16.1)
Single
Educational attainment
0.767
High school/GED
16 (8.3)
1 (5.9)
15 (8.5)
Some college
135 (69.9) 42 (21.8)
11 (64.7) 5 (29.4)
124 (70.5) 37 (21.0)
College graduate
Body mass index
0.154
Normal weight
32 (16.9) 101 (53.4) 56 (29.6)
5 (31.3) 9 (56.3) 2 (12.5)
27 (15.6) 92 (53.2) 54 (31.2)
Overweight
Obese
Influenza shot in the past 12 months
0.027
Yes No
35 (18.9) 150 (81.1)
0 (0.0)
35 (21.0) 132 (79.0)
18 (100.0)
Smoking status
0.543
Current smoker Former smoker Never smokers
0 (0.0) 8 (4.6)
0 (0.0) 1 (6.3)
0 (0.0) 7 (4.4)
167 (95.4)
15 (93.8)
152 (95.6)
Career firefighter tenure
0.419
Years±SD
15.9 ± 9.2
14.1 ± 8.1
16.0 ± 9.3
Time at current department
0.732
Years±SD
15.3 ± 9.1
14.6 ± 8.4
15.3 ± 9.1
Current rank
0.129
Firefighter/paramedic/EMT
79 (40.7) 30 (15.5)
4 (22.2) 5 (27.8) 1 (5.6) 7 (38.9) 0 (0.0) 1 (5.6)
75 (42.6) 25 (14.2)
Driver/operator
Inspector/fire investigator
7 (3.6)
6 (3.4)
Lieutenant
47 (24.2) 18 (9.3) 13 (6.7)
40 (22.7) 18 (10.2) 12 (6.8)
Captain
Battalion/deputy/division chief Any symptoms in the past 2 weeks
0.064
Yes No
18 (9.0)
4 (22.2) 14 (77.8)
14 (7.7)
181 (91.0)
167 (92.3)
Days with symptoms since onset
0.005
Average days±SD
5.7 ± 4.3
11.7 ± 2.5
4.6 ± 3.7
COVID-19 case contact past 2 weeks at work
0.653
Yes No
93 (51.7) 43 (23.9) 44 (24.4)
7 (43.8) 5 (31.3) 4 (25.0)
86 (52.4) 38 (23.2) 40 (24.4)
Not sure
continued
Caban-Martinez AJ, et al . Occup Environ Med 2020; 77 :857–861. doi:10.1136/oemed-2020-106676
858
Workplace
Table 1 continued
COVID-19 antibody test result
Total sample N N (%)*
Positive (IgG, IgM or IgG/IgM) n (%)*
Negative n (%)*
Characteristics
P value
Average COVID-19 case contacts
0.022
Average cases±SD
7.73 ± 6.3
13.3 ± 4.8
7.31 ± 6.2
Average time spent with COVID-19 positive person
0.025
Less than 5 min
15 (17.0) 66 (75.0)
6 (42.9) 8 (57.1) 0 (0.0)
9 (12.2) 58 (78.4)
5–30 min
Greater than 30 min
7 (8.0)
7 (9.5)
Used any PPE during COVID-19 encounter
0.383
Yes, any PPE No PPE use
87 (93.5)
6 (85.7) 1 (14.3)
81 (94.2)
6 (6.5)
5 (5.8)
Number of PPE items used with COVID-19 positive person
0.79
Average count PPE used (one to six items) 3.3 ± 1.4 *Differences in subtotal population sample due to item non-response or missing. Case contact=when a firefighter was within 6 feet of an infected person (laboratory-confirmed or probable COVID-19 patients) for at least 15 min; COVID-19 positive person indicates an individual with laboratory-confirmed COVID-19 test; PPE items included gloves, double gloves, N-95 respirator, fluid resistant sleeves, eye protection and gown. EMT, Emergency Medical Technician; GED, General Educational Development; Ig, immunoglobulin; PPE, personal protective equipment. 3.3 ± 1.4 3.1 ± 1.6
drove through the antibody testing fire station in their personal vehicles or on-duty engine/rescue trucks for SARS-CoV-2 anti- body testing. Off-duty firefighters/paramedics could wait in the fire station parking lot for 10–15 min to receive the results of their antibody test, while on-duty personnel were instructed to return to their fire station immediately. The fire department infec- tion control officer (ICO) followed up directly with any on-duty firefighter personnel who tested positive for SARS-CoV-2 anti- bodies. The ICO immediately quarantined the firefighter/para- medic, conducted reflex nasal swab (reverse transcription PCR (RT-PCR)) testing and closely monitored coworker firefighters at the fire station as part of the comprehensive F-TRACE project. A total of three firefighters/paramedics did not participate in the surveillance project, of which two were out of the geographic area because of scheduled vacation and one declined to partici- pate for religious reasons (response rate=98.6%). Study survey measures and administration Firefighters/paramedics were asked to complete two web-based survey instruments (ie, an intake form and a COVID-19 expo- sure form) prior to their scheduled antibody testing day. Survey instruments were administered to firefighters/paramedics via an email link sent by fire department leadership using REDCap, a secure web-based application for building and managing complex online surveys and databases. 10 The intake form consisted of a 30-item questionnaire assessing sociodemographic (ie, age, sex, race, ethnicity, educational attainment, marital status, height/ weight and contact information) and work characteristics (ie, station assignment, shift schedule, firefighter tenure, rank, current job tasks, number of fire/EMS calls, second job and mili- tary experience) adapted from questions on federal surveys. 11 The COVID-19 exposure form is comprised of 29 items assessing COVID-19 firefighter/paramedic symptoms, prior COVID-19 testing, COVID-19 contacts, smoking status, receipt of influenza shot in prior 12 months, as well as a series of ques- tions on exposure risk assessing the frequency and duration of COVID-19 patient exposures, PPE use and firefighter coworker contacts adapted from the CDC COVID-19 questionnaire. 12 We assessed the type of PPE used by the firefighter by asking, “Were you using any protective equipment when you came into contact with possible COVID-19 person? Choose all that apply.” Response options included: “gloves, double gloves, N-95 respi- rator, fluid resistant sleeves, eye protection, and gown”. A case
regarding the prevalence of coronavirus in the workforce. Strat- egies that limit the spread of the SARS-CoV-2 virus within their workforce and tools that provide near real-time decision-making on firefighter/paramedics return to work algorithms and infec- tion control strategies are needed. Serological antibody tests, despite their limitations, are critical tools for assessments of SARS-CoV-2 exposure, infection and potential immunity. 4 5 Current testing for the SARS-CoV-2 virus largely depends on labour-intensive molecular techniques that can often be delayed by days, limiting their utility in return to work algorithms for a fast-paced emergency responder workforce. 6 7 Recent studies have documented that asymptomatic individuals might contribute to SARS-CoV-2 transmission, further compli- cating efforts to limit the spread of the virus. 8 9 As part of a complementary and broader comprehensive COVID-19 medical surveillance programme, reliable antibody detection assays would enable more accurate estimates of SARS-CoV-2 prev- alence and incidence in the first responder workforce. A joint collaborative partnership between city government, fire depart- ment, local union and an academic medical centre supported the implementation of the Firefighter Tracking, Resources, and Assessment of COVID-19 Epidemiology (F-TRACE) project, supporting the coordination, tracking and educational resources of COVID-19 contact, presumptive and confirmed cases among fire department personnel. In the present study, we estimate the point seroprevalence of SARS-CoV-2 antibodies among frontline firefighter/paramedics of a South Florida fire department located in the epicentre of a State outbreak. Secondary data analysis of cross-sectional information collected as part of the fire department’s F-TRACE project was used to esti- mate the point seroprevalence of SARS-CoV-2 antibodies among frontline firefighters/paramedics collected over a 2-day period, 16–17 April 2020. Firefighters/paramedics of a US fire depart- ment in Florida were invited by department and local union leadership to voluntarily consent to participate in a one-time surveillance assessment for SARS-CoV-2 antibodies. Fire depart- ment personnel were emailed a survey link assessing COVID-19 symptoms and work exposure characteristics the day prior to the scheduled drive-through antibody testing at a designated fire station. Off-duty and on-duty firefighter/paramedic personnel METHODS Study design, participants and recruitment
Caban-Martinez AJ, et al . Occup Environ Med 2020; 77 :857–861. doi:10.1136/oemed-2020-106676
859
Workplace contact was defined as a firefighter who was within 6 feet of an infected person (laboratory-confirmed or probable COVID-19 patients) for at least 15 min; the definition is consistent with the US CDC guidelines. SARS-CoV-2 antibody test administration A rapid immunoglobulin (Ig)M-IgG combined point-of-care (POC) lateral flow immunoassay (BioMedomics, Morrisville, New Carolina, USA) was used for assessment of SARS-CoV-2 antibodies in participating firefighters/paramedics. 13 The sensi- tivity and specificity of the COVID-19 antibody assay were estimated to be 88.66% and 90.63%, respectively, based on the results for 397 infected cases and 128 non-SARS-CoV-2 infection patients in Wuhan, China. 14 The positive predictive value (PPV) of the test is estimated to be 33.2% and the negative predictive value is 99.3%. On testing day, on-duty and off-duty firefighters/ paramedics drove through a structured drive-through lane at the fire station/training facility. Firefighters/paramedics rolled down their window, were approached by the F-TRACE team gowned in PPE for an initial index finger swab with rubbing alcohol, followed by a quick lancet finger puncture to allow for two drops of blood to be placed in the cassette sample well. F-TRACE team members added two drops of buffer reagent to the cassette sample well and waited 10 min for the test to complete prior to reading the results. Data analysis We calculated descriptive statistics for continuous variables, expressed as means with its SD, and for categorical variables, expressed as frequency and percent of the sample. We examined the main outcome of testing positive (combined IgG only, IgM only and IgG/IgM) by sociodemographic and work character- istics, by COVID-19 contacts and COVID-19 symptoms. For categorical data, we conducted Fisher’s exact test to compare groups. Student’s t-test was used to compare the mean days of symptom onset, firefighter tenure, time in fire department, average number of COVID-19 case contacts, average time spent with COVID-19 cases and the number of PPE items used with COVID-19 case between firefighter who tested positive versus negative. P values less than 0.05 were considered statistically significant. We performed all data management and statistical analyses using SPSS V.26 for Windows (IBM). RESULTS Among the 203 firefighters/paramedics that participated in the F-TRACE project, 18 firefighters/paramedics (8.9%) tested positive for SARS-CoV-2 antibodies, of which 8 firefighters/ paramedics (3.9%) were IgG positive only, 8 (3.9%) were IgM positive only and 2 (0.1%) were IgG/IgM positive (table 1). None of the antibody positive firefighters/paramedics reported receipt of the annual influenza vaccine compared with fire- fighters/paramedics who tested negative for SARS-CoV-2 anti- bodies (0.0% vs 21.0%; p=0.027). Although not significant, the proportion of firefighters/paramedics who reported symptoms in the 2 weeks prior to antibody testing was higher for those who tested antibody positive compared with firefighters/paramedics who were antibody negative (22.2% vs 7.7%; p=0.064). The average number of COVID-19 case contacts was significantly higher (13.3±4.8 case contacts vs 7.31±4.8 contacts; p=0.022) among firefighters/paramedics who were SARS-CoV-2 antibody positive compared with firefighters who tested negative for antibodies.
DISCUSSION As a component of an overall medical surveillance programme, we found variation in the seroprevalence of SARS-CoV-2 antibodies among frontline firefighters/para- medics of a moderately sized US fire department. Approxi- mately 4% of the participating firefighters/paramedics tested positive for either IgM or IgG/IgM SARS-CoV-2 antibodies, indicating recent infection from the time of immunoassay antibody testing. These findings provided timely and useful information on decision to quarantine and further evaluation through reflex RT-PCR nasal swabs. Nonetheless, caution at interpreting the results of the antibody testing is warranted. When the prevalence of COVID-19 is based on serology testing (ie, all antibodies including IgM only, IgG only and combined IgM/IgG) and the prevalence of COVID-19 is esti- mated to be low (eg, 5% within the workforce), the risk of false positives can be elevated. For example, if the COVID-19 serological test has 90% specificity, we estimate that its PPV will be 32.1%, thus nearly 70% of positive results will likely be false. At this same disease prevalence (~5% of the work- force), a test with 95% specificity will lead to a 50% chance that a positive result is incorrect. Similarly, it is possible that a positive result on COVID-19 antibody serology test can be due to cross-reactivity with other viruses. Different assays use antigens from different parts of SARS-CoV-2, and some combine IgM and IgG, therefore different levels of cross- reactivity with other coronavirus antibodies are possible. 15 At the time of this pilot study (April 2020), the BioMedomics COVID-19 assay was used by our team under the Emer- gency Use Authorisation (EUA) authority of the US Federal Drug Administration for research and community surveil- lance to estimate COVID-19 infectivity within the firefighter workforce. We found that firefighter/paramedics who tested posi- tive were significantly more likely to have greater number of contacts with COVID-19 positive patients as well as spend less time (less than 5 min) with COVID-19 positive patients compared with firefighter/paramedics who tested SARS-CoV-2 antibody negative. Among all firefighters/para- medics who tested SARS-CoV-2 antibody positive, none had reported receipt of the annual influenza vaccine in the 12 months prior to antibody testing. It may be possible that those firefighters/paramedics who tested positive engage in riskier behaviour (ie, inconsistent use of PPE) that could lead to greater risk of exposure. It is possible that an individual’s vaccination behaviour can provide insight into their overall risk tolerance and work-related safety practices. For example, community-based studies evaluating risky sexual behaviour among young adults who were vaccinated against human papillomavirus showed they engaged in less risky behaviours such as being less likely to not use a condom and drink two or more times per week. 16 In our study, we found that PPE use (while not significant) was lower for firefighters who tested positive versus those who tested COVID-19 antibody negative (85.7% vs 94.2%)—more than a twofold difference. A recent systematic review examining clustering and co-occurrence of multiple risk behaviours (ie, drinking, physical activity, diet and so on) found the strongest associations by occupation (up to fourfold increased odds) and by educational attainment. 17 Furthermore, in other occupational groups like the construc- tion workforce, 18 health behaviours have been linked to safety perceptions where obese construction workers with low phys- ical activity were less concerned about job-related injuries.
Caban-Martinez AJ, et al . Occup Environ Med 2020; 77 :857–861. doi:10.1136/oemed-2020-106676
860
Workplace
Understanding how vaccination practices impact PPE use in first responders, a US occupational group with historically low vaccination rates, 19 can shed insight to potential strategies to improve PPE use and COVID-19 control strategies. Nonethe- less, the relatively small sample size warrants repeated data collection with other fire departments. Further longitudinal research is needed to further investigate long-term immunity to the SARS-CoV-2 virus among first responders, particularly how the use of PPE mitigates rates of infection and how sero- logical antibody testing can inform return to work strategies. Twitter Alberto J Caban-Martinez @DrCabanMartinez and Erin Kobetz @http:// www.twitter.com/K obetzErin Acknowledgements We would like to acknowledge the fire department, the firefighters and local union of the International Association of Fire Firefighters (IAFF) for collaborating in the F-TRACE medical surveillance project. Contributors AJC-M, SBI and EK contributed to the conceptualisation and the design of the work, statistical analysis and interpretation of data and final drafting of the document. AJC-M, KS, MG, AB and PLF contributed to the statistical analysis and interpretation of the data, drafting the document and final approval. AJC-M, NS-S, KS, PLF, AB, MG, SBI and EK contributed to the interpretation of the data, drafting the document and final approval. All authors agree to be held accountable for all aspects of the work related to its accuracy and integrity. Funding Support for this research is, in part, by the State of Florida appropriation # 2382A for Firefighter Cancer Initiative (principal investigator (PI): EK) to the University of Miami (UM) Sylvester Comprehensive Cancer Center; the Federal Emergency Management Administration (FEMA) Grant # EMW-2017-FP-00860 (PI: AJC-M); the State of Florida appropriation #62 for the Medical Training and Simulation Laboratory to the University of Miami Gordon Center for Simulation and Innovation in Medical Education (PI: SBI); and by the National Cancer Institute of the National Institutes of Health under Award Number P30CA240139. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or Federal Emergency Management Agency of the Department of Homeland Security. Competing interests None declared. Patient consent for publication Not required. Ethics approval Review of the study protocol and approval was by the University of Miami Institutional Review Board (#20200537). Provenance and peer review Not commissioned; externally peer reviewed. Data availability statement Data are available on reasonable request. Data from this study are available on request by sending an email message to the corresponding author, Dr Alberto Caban-Martinez (a caban@med.m iami.edu). This article is made freely available for use in accordance with BMJ’s website terms and conditions for the duration of the covid-19 pandemic or until otherwise
determined by BMJ. You may use, download and print the article for any lawful, non-commercial purpose (including text and data mining) provided that all copyright notices and trade marks are retained. ORCID iD Alberto J Caban-Martinez http://orcid.org/0 000-0002-5 960-1308 REFERENCES 1 Jernigan DB, CDC COVID-19 Response Team. Update: Public Health Response to the Coronavirus Disease 2019 Outbreak - United States, February 24, 2020. MMWR Morb Mortal Wkly Rep 2020;69:216–9. 2 Gamio L. The workers who face the greatest coronavirus risk. New York Times 2020. 3 Spina S, Marrazzo F, Migliari M, et al . The response of Milan’s emergency medical system to the COVID-19 outbreak in Italy. Lancet 2020;395:e49–50. 4 Goudsmit J. The paramount importance of serological surveys of SARS-CoV-2 infection and immunity. Eur J Epidemiol 2020;35:331–3. 5 Long Q-X, Liu B-Z, Deng H-J, et al . Antibody responses to SARS-CoV-2 in patients with COVID-19. Nat Med 2020;26:845–8. 6 Patel R, Babady E, Theel ES, et al . Report from the American Society for microbiology COVID-19 international Summit, 23 March 2020: value of diagnostic testing for SARS–CoV-2/COVID-19. MBio 2020;11. 7 Kimball A. Asymptomatic and presymptomatic SARS-CoV-2 infections in residents of a long-term care skilled nursing facility—King County, Washington, March 2020. MMWR Morbidity and mortality weekly report 2020:69. 8 Bai Y, Yao L, Wei T, et al . Presumed asymptomatic carrier transmission of COVID-19. JAMA 2020;323:1406–7. 9 Ling Z, Xu X, Gan Q, et al . Asymptomatic SARS-CoV-2 infected patients with persistent negative CT findings. Eur J Radiol 2020;126:108956. 10 Harris PA, Taylor R, Thielke R, et al . Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42:377–81. 11 NCfH S. National health interview survey questionnaire . Hyattsville, MD: National Center for Health Statistics, 2000. 12 Burke RM. Active monitoring of persons exposed to patients with confirmed COVID-19—United states, January–February 2020. MMWR Morbidity and mortality weekly report 2020:69. 13 Li Z, Yi Y, Luo X, et al . Development and clinical application of a rapid IgM‐IgG combined antibody test for SARS‐CoV‐2 infection diagnosis. J Med Virol 2020. 14 Li Z, Yi Y, Luo X, et al . Development and clinical application of a rapid IgM‐IgG combined antibody test for SARS‐CoV‐2 infection diagnosis 2020. 15 Petherick A. Developing antibody tests for SARS-CoV-2 2020;395:1101–2. 16 Brouwer AF, Delinger RL, Eisenberg MC, et al . Hpv vaccination has not increased sexual activity or accelerated sexual debut in a college-aged cohort of men and women. BMC Public Health 2019;19:821. 17 Meader N, King K, Moe-Byrne T, et al . A systematic review on the clustering and co- occurrence of multiple risk behaviours. BMC Public Health 2016;16:1–9. 18 Strickland JR, Wagan S, Dale AM, et al . Prevalence and perception of risky health behaviors among construction workers. J Occup Environ Med 2017;59:673–8. 19 Caban-Martinez AJ, Lee DJ, Davila EP, et al . Sustained low influenza vaccination rates in US healthcare workers. Prev Med 2010;50:210–2.
Caban-Martinez AJ, et al . Occup Environ Med 2020; 77 :857–861. doi:10.1136/oemed-2020-106676
861
VIEWPOINT
Vulnerability of the medical product supply chain: the wake-up call of COVID-19
Fiona A Miller , 1 Steven B Young, 2 Mark Dobrow, 1 Kaveh G Shojania 3
1 Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Ontario, Canada 2 School of Environment, Enterprise and Development, University of Waterloo, Waterloo, Ontario, Canada 3 Department of Medicine, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada Correspondence to Dr Fiona A Miller, Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Canada; fiona.miller@u toronto.ca
acquire. Many shortages seen during the COVID-19 crisis clearly affect safety, and they can exacerbate widespread problems with equity. Just as critical incidents afford the opportunity to identify not just obvious active errors but also latent safety prob- lems, 6 crises such as COVID-19 expose general supply chain weaknesses. In fact, product shortages often exhibit the combination of active and latent errors (or ‘system problems’) seen in investiga- tions of critical safety incidents. Hurricane Maria in 2017, for example, converted a chronic shortage of sterile saline solu- tions (for intravenous administration) into an acute shortage when manufac- turing capacity concentrated in Puerto Rico was damaged. 7 The current ampli- fied risk of generic drug shortages follows from several decades of chronic shortage associated with fewer firms and concen- trated sites of production. 8 These crit- ical incidents are unlikely to abate, given the continued threat of future infectious disease outbreaks, 9 and the accelerating climate crisis, which will increase extreme weather, violent conflicts and other events that provoke acute shortages. 10 The vulnerability of medical product supply—and its converse, resilience—has historically attracted little interest from clinicians, healthcare executives or those engaged in improving healthcare quality. Yet, as the COVID-19 pandemic has made obvious to even the casual observer, product shortages affect clinical practice, organisational performance and patient outcomes. In this article, we outline what is known from the extensive literature on supply chain resilience and medical product shortage and use examples from both healthcare and non-healthcare
INTRODUCTION The COVID-19 pandemic has brought the long-standing vulnerability of the medical product supply chain into sharp focus. Global shortages of medical prod- ucts accompanied the global spread of the disease, joined by high prices, the proliferation of suspect dealers and dramatic interventions by governments, philanthropy and industry in oftentimes- unsuccessful attempts to secure solutions. Much attention has focused on personal protective equipment (PPE). But reported shortages have extended much further—to testing supplies, dial- ysis materials, pharmaceuticals and a wide range of commodities essential for daily care delivery—both for patients with and without COVID-19. 1 2 PPE shortages have received partic- ular attention because they endanger the healthcare workforce. 3 But all product shortages endanger patients due to delays in care, rationing or denial of care, the use of substandard products, or heightened risk of error when using replacement products—risks that extend to increased mortality. 4 Medical product shortages threaten the goal to deliver the right care to the right person at the right time—and have done so for decades. 5 The COVID-19 pandemic has highlighted more than ever that these systemic risks can no longer be ignored. It may also mean that new solu- tions have become more possible. Why care about medical supply chains? Unexpected shortages of medical prod- ucts do not fit neatly into any single quality domain, but can affect all of them. We cannot provide effective, efficient or timely care when medicines and other supplies required for crucial elements of care become difficult or impossible to
Received 2 August 2020 Revised 17 October 2020 Accepted 22 October 2020 Published Online First 2 November 2020
© Author(s) (or their employer(s)) 2021. No
To cite: Miller FA, Young SB, Dobrow M, et al . BMJ Qual Saf 2021; 30 :331–335. commercial re-use. See rights and permissions. Published by BMJ.
331
Miller FA, et al . BMJ Qual Saf 2021; 30 :331–335. doi:10.1136/bmjqs-2020-012133
Viewpoint
Table 1 Manufacturing problems as causes of supply chain disruptions Risks to manufacture Non-medical examples
Medical examples
► ► Earthquake and tsunami in Japan, and flooding in Thailand, disrupts automotive and electronics industries in 2011 35
► ► Acute sterile saline shortage after hurricane damage to manufacturing plants, 2017 7 ► ► Risk of medical gloves shortage due to COVID-19 lockdown in Malaysia, the home of ~65% of the global supply 36 ► ► Critical shortage of propofol (2009–2010) as two of three suppliers left US market 38 ► ► Shortages of nine childhood vaccines, 2000–2005, due to problems with the limited numbers of suppliers 39 ► ► Disruptions in supply of raw or bulk materials responsible for drug shortages 33 ► ► Lack of meltblown, non-woven polypropylene for the production of surgical masks and N95 respirators 41
Geographically concentrated manufacturing : Production concentrated in one or few locations, such that a localised disruptive event (natural or political) risks major disruption to product manufacturing 13 Limited numbers of manufacturers : Few manufacturers, such that events affecting a single firm (eg, disruptions or decisions) risks major disruption to product manufacturing 17 Scarcity of critical inputs : Resource inputs or parts whose scarcity risks major disruption due to non-substitutability of resources 14 or tightly-coupled production arrangements (just-in-time, short-cycle manufacturing) 17
► ► Fire at Philip’s semiconductor plant in 2000 disrupts Ericson’s sole-source of chips for mobile phone production 37
► ► Multiyear delays in production of Boeing Dreamliner due to shortage of aerospace fasteners 40
industries to illustrate key vulnerabilities. As well, we offer examples of how these vulnerabilities have been exposed by the COVID-19 crisis. We consider some of the reactive adaptations forced on clinicians and administrators, most notably by PPE shortages, and identify several common failures of pandemic planning. Because the vulnerability of medical product supply long pre-dates the pandemic, we also highlight the need for remedies that extend beyond pandemic response capacity—including from bold experimen- tation at the front line and by governments. Such reforms are contested and their prospects uncertain— no problem of this nature is amenable to easy solu- tions. Yet successfully addressing any quality problem begins with understanding its contributing factors. Thus, while we identify some promising reform direc- tions, our main goal lies in outlining current knowl- edge about the factors contributing to supply chain disruptions and highlighting the need for broad and sustained engagement with the challenge of resilient medical product supply.
Understanding supply chain vulnerability for medical products Quite a bit is known about what makes complex, contemporary global supply chains so vulnerable. Because no single, widely used framework for charac- terising supply chain vulnerabilities exists, particularly from the perspective of the supply user, we discuss examples in two broad categories—threats to product manufacture (table 1) and threats to local availability (table 2). Manufacturing problems as risks to product supply chains The consumers of medical—or other—products are often unaware of vulnerabilities in supply until short- ages occur—when manufacture of the product has ceased or no longer occurs in sufficient quantities (table 1). Either event can occur when production is concentrated among few firms, which may elect to exit the market, or in few places, such that political or geophysical events disrupting local manufacture have global consequences. 11 For example, severe flooding in
Table 2 Transportation, regulation and supply chain management as threats to product availability Risks to availability Non-medical examples
Medical examples
► ► Oil tanker stuck in Suez Canal blocked ships carrying PlayStation II to consumers for 2004 Christmas season 42 ► ► Intentional contamination of milk with melamine identified in 2008, leading to health harms and food recalls 43
► ► Export bans, authorisations and related restrictions on critical supplies 19 ► ► Grounding of commercial flights challenge global freight movement 19 ► ► Heparin recall in 2008 after deaths and adverse events due to manufacturing plant adulteration 44 ► ► Multiple recalls of medicines containing valsartan in 2018 due to contamination risk 45 ► ► Risks from single vendor contracts (eg, Sprint Fidelis defibrillator lead recall of 2007 47 ) ► ► Poor contract management leading to drug shortages 48
Restrictions on the mobility of goods from the place of production to the place of consumption due to limitations on cross-border flows or transportation impedances 17 Restrictions on quality and availability in specific markets due to regulatory compliance failures : Relevant to industries regulated for public or consumer safety, for example, medical products, aviation, automotive, telecommunication 18
► ► Retail sourcing emphasis on large-scale producers lead to severe vegetable shortages in Europe, winter 2016–2017 46
Short-termism in sourcing activities : Sourcing activities that do not prioritise long-term supply reliability (eg, focus on upfront costs, sole sourcing, poor contract management) 8
332
Miller FA, et al . BMJ Qual Saf 2021; 30 :331–335. doi:10.1136/bmjqs-2020-012133
Viewpoint
Thailand in 2011 puts factories producing 43% of the world’s hard disk drives underwater, reducing produc- tion of personal computers for months and reducing global industrial output by 2.5%. 12 The ability to manufacture products is also threat- ened by shortages of inputs, whether of raw resources or component parts. Input shortages may result from tightly coupled production arrangements, which limit inventory or increase complexity. 13 The severity of a shortage depends on the input’s ‘criticality’, including its importance and the availability of potential substi- tutes or other mitigation strategies. 14 Production vulnerabilities during the COVID-19 pandemic partly reflect drastically elevated demand, such as for PPE. 3 But production vulnerabilities can arise even with normal demand, especially when production is geographically concentrated, as with medical gloves and generic drug manufacture in India. 15 Moreover, the lean nature of the medical products industry is likely to challenge the capacity to produce sufficient quantities of any therapeutics that are shown to be effective in treating the disease. 16 Other threats to local product availability Even when manufacture occurs normally, multiple vulnerabilities can reduce availability of the right product in the right place (table 2). Lengthy trans- portation routes can prove fragile in times of need. 17 Additional challenges arise where companies fail to comply with the regulatory requirements that aim to reduce the risk of flawed products (eg, from the US Food and Drugs Administration or European Medi- cines Agency). 18 Conversely, inadequate regulatory arrangements can mean that available products are unsafe or product shortages are unexpected. 8 Vulnerabilities affecting the availability of products during the COVID-19 pandemic have included move- ment limitations due to the reduction of international transportation capacity as well as the imposition of border controls and export restrictions, as countries prioritised the needs of their own citizens over those of international clients. 19 Further, those charged with sourcing products have been exposed as ill-prepared. Few had sourced supplies with a view to long-term resilience (eg, by favouring reusable products or domestic suppliers), managed PPE as a critical sector requiring vendor monitoring and risk management, or maintained supplies sufficient to avoid stockouts and shortages, including of goods such as generic drugs, for which demand has not been markedly elevated. 20 21 Reactive and proactive solutions for medical product shortage Shortages during the COVID-19 crisis have forced reactive adaptations. For clinicians, this has included shifts in normal standards of PPE use, such as extended use (eg, wearing the same PPE for encounters with different patients), reuse after sterilisation, alternative
products (eg, positive pressure airflow helmets rather than N95 respirators) and even non-use. 22 The health and safety risks such reactive changes have created remain unclear. But shifts in policy on PPE use have challenged healthcare professionals’ trust in system leaders and governments. 23 For healthcare adminis- trators, product shortages make sourcing efforts much more complex. Many have had to deal with unknown and sometimes fraudulent suppliers or compete with other care delivery organisations for needed supplies, with concerning implications for care quality and equi- table access. 24 Virtually all have faced increased costs as well as operational, legal or reputational risks. 25 To an important extent, reactive efforts by clini- cians and administrators have been made necessary by pandemic planning policy failures. Many health systems had not built or adequately sustained national or regional stockpiles. Many agencies tasked with coordinating national or regional joint procurement efforts lacked capacity. Much of the information infrastructure needed to fairly allocate supplies to users across countries or regions has proven insuf- ficient. 26 27 And the pandemic has again exposed the weakness of market access regulation with respect to manufacturers’ obligations to notify about, or provide assurances of, reliable and quality supply. 28 Yet while necessary, more effective pandemic plan- ning will not be a sufficient response. The underlying causes of supply chain vulnerability are not specific to the pandemic. They include the economic and regu- latory arbitrage that lead brand name manufacturers to restrict generic competition, generic manufacturers to discontinue (or underinvest in) less profitable product lines, product supply chains to metastasise into complex networks of global manufacture and transport, and healthcare buyers to undervalue resil- ient supply. Importantly, the pandemic has increased experimen- tation with proactive solutions, some of which take aim at these systemic challenges. The most promising, in our view, include increased interest in medical prod- ucts that can be reused or repurposed, alongside the free and open-source hardware principles that support distributed and locally responsive refurbishment and manufacturing. Such experimentation has been most notable for PPE, often driven by front-line clinicians. 29 For more complex medical products, such as therapeu- tics or vaccines, governments and international agen- cies have experimented with ways to ensure adequate manufacturing capacity. This includes coordinated global strategies, such as advance market commitments for novel vaccines, 30 and the turn to enhanced domestic manufacturing and reduced dependence on limited suppliers of critical inputs for pharmaceuticals. 31 Less prominent but not absent have been challenges to the intellectual property rights that protect companies’ monopoly control and high prices, through efforts to create global technology access pools and national
333
Miller FA, et al . BMJ Qual Saf 2021; 30 :331–335. doi:10.1136/bmjqs-2020-012133
Viewpoint threats of compulsory licensing. 16 30 Such moves are not uncontested, nor is their effectiveness assured. 32 Yet they represent an important expansion of oppor- tunity for reform—an opportunity whose potential will rest, in part, on the collective engagement of the healthcare community CONCLUSION Many of the policy and system issues that bear on medical product supply are not traditional areas of concern for the healthcare quality community. Quality improvement teams may take aim at medication administration errors due to poor labelling but not the shortage-induced use of unfamiliar products that contribute to such errors. They may target delays in care due to poor scheduling but not delays that arise when needed products are simply not available. Quality improvement teams may not lack interest, but they will often lack leverage. Individual clinicians and care organisations can anticipate shortages and mitigate their harms once they arise (eg, the Amer- ican Society of Health-System Pharmacists’ Guide- lines on Managing Drug Product Shortages), 33 but their capacity to prevent such shortages is inherently limited. Resilient sourcing strategies offer some reme- dies—sourcing from multiple vendors, securing local supply and maintaining inventory. But demand side strategies can only do so much. As with the persistent neglect of human factor issues in medical device design, many solutions necessitate ‘controlling the supply side’ at macroscale. 34 The pandemic may have made some of these controls more possible, but there are few easy solutions to systemic patient safety chal- lenges. As with addressing any quality problem, the first step consists of recognising and understanding the contributing factors and system issues. The next lies with assuming a shared responsibility in developing effective solutions. Contributors FM and KGS conceived the idea for the paper. FM drafted the manuscript. All authors contributed to manuscript revisions. Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Competing interests None declared. Patient consent for publication Not required. Provenance and peer review Not commissioned; externally peer reviewed. This article is made freely available for use in accordance with BMJ’s website terms and conditions for the duration of the covid-19 pandemic or until otherwise determined by BMJ. You may use, download and print the article for any lawful, non-commercial purpose (including text and data mining) provided that all copyright notices and trade marks are retained. ORCID iDs Fiona A Miller http://orcid.org/0000-0003-4953-6255 Kaveh G Shojania http://orcid.org/0000-0002-9942-0130
REFERENCES 1 Office of Inspector General of the Department of Health and Human Services US. Hospital experiences responding to the COVID-19 pandemic: results of a national pulse survey March 23-27, 2020 . Washington, DC: US Department of Health and Human Services, Office of Inspector General, 2020. 2 Ranney ML, Griffeth V, Jha AK. Critical Supply Shortages - The Need for Ventilators and Personal Protective Equipment during the Covid-19 Pandemic. N Engl J Med 2020;382:e41. 3 WHO. Shortage of personal protective equipment endangering health workers worldwide . Geneva: World Health Organization, 2020. 4 Phuong JM, Penm J, Chaar B, et al . The impacts of medication shortages on patient outcomes: a scoping review. PLoS One 2019;14:e0215837. 5 Tucker EL, Cao Y, Fox ER, et al . The drug shortage era: a scoping review of the literature 2001–2019. Clinical Pharmacology & Therapeutics 2020:n/a(n/a). 6 Trbovich P, Shojania KG. Root-cause analysis: swatting at mosquitoes versus draining the swamp. BMJ Quality & Safety 2017;26:350–3. 7 Sacks CA, Kesselheim AS, Fralick M. The shortage of normal saline in the wake of Hurricane Maria. JAMA Internal Medicine 2018;178:885–6. 8 Food, Drug Administration US. Drug Shortages: Root Causes and Potential Solutions: A Report by the Drug Shortages Task Force: US Food & Drug Administration. 2019 [updated 2020. 9 Karesh WB, Dobson A, Lloyd-Smith JO, et al . Ecology of zoonoses: natural and unnatural histories. Lancet 2012;380:1936–45. 10 Andreoni V, Miola A. Climate vulnerability of the Supply- Chain: literature and methodological review. Eur – scientific and technical research series. Luxembourg: Publications Office of the European Union 2014;118. 11 Woodcock J, Wosinska M. Economic and technological drivers of generic sterile injectable drug shortages. Clinical Pharmacology & Therapeutics 2013;93:170–6. 12 Haraguchi M, Lall U. Flood risks and impacts: A case study of Thailand’s floods in 2011 and research questions for supply chain decision making. International Journal of Disaster Risk Reduction 2015;14:256–72 https://doi.org/ 13 Ponomarov SY, Holcomb MC. Understanding the concept of supply chain resilience. The International Journal of Logistics Management 2009;20:124–43. 14 Cimprich A, Bach V, Helbig C, et al . Raw material criticality assessment as a complement to environmental life cycle assessment: examining methods for product-level supply risk assessment. Journal of Industrial Ecology 2019;23:1226–36. 15 Guerin P, Singh-Phulgenda S, Strub-Wourgaft N. The consequence of COVID-19 on the global supply of medical products: Why Indian generics matter for the world? [version 1; peer review: 1 approved]. F1000Research 2020;9(225. 16 OECD. Treatments and a vaccine for COVID-19: the need for coordinating policies on R&D, manufacturing and access . Paris: Organisation for Economic Cooperation and Development, 2020. 17 Hohenstein N-O, Feisel E, Hartmann E, et al . Research on the phenomenon of supply chain resilience: a systematic review and paths for further investigation. International Journal of Physical Distribution & Logistics Management 2015;45:90–117.
334
Miller FA, et al . BMJ Qual Saf 2021; 30 :331–335. doi:10.1136/bmjqs-2020-012133
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 21Powered by FlippingBook