Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Kroth PJ, Morioka-Douglas N, Veres S, et al. Association of Electronic Health Record Design and Use Factors With Clinician Stress and Burnout. JAMA Netw Open. 2019;2(8):e199609. doi:10.1001/jamanetworkopen.2019.9609
Which electronic health record (EHR) design and use factors are associated with clinician stress and burnout?
In this survey study of 282 clinicians, clinician stress and burnout were associated with 7 EHR design and use factors. These 7 plus 2 other design and use factors collectively accounted for a modest amount of the variance in stress (12.5%) and burnout (6.8%); models incorporating other work conditions (such as chaotic work atmosphere and workload control) accounted for considerably more of the variance in stress (58.1%) and burnout (36.2%).
While EHR design and use factors may appropriately be targeted by health systems and EHR designers to address stress and burnout, other non-EHR issues, especially clinician work conditions, appear to play a substantial role in adverse clinician outcomes.
Many believe a major cause of the epidemic of clinician burnout is poorly designed electronic health records (EHRs).
To determine which EHR design and use factors are associated with clinician stress and burnout and to identify other sources that contribute to this problem.
Design, Setting, and Participants
This survey study of 282 ambulatory primary care and subspecialty clinicians from 3 institutions measured stress and burnout, opinions on EHR design and use factors, and helpful coping strategies. Linear and logistic regressions were used to estimate associations of work conditions with stress on a continuous scale and burnout as a binary outcome from an ordered categorical scale. The survey was conducted between August 2016 and July 2017, with data analyzed from January 2019 to May 2019.
Main Outcomes and Measures
Clinician stress and burnout as measured with validated questions, the EHR design and use factors identified by clinicians as most associated with stress and burnout, and measures of clinician working conditions.
Of 640 clinicians, 282 (44.1%) responded. Of these, 241 (85.5%) were physicians, 160 (56.7%) were women, and 193 (68.4%) worked in primary care. The most prevalent concerns about EHR design and use were excessive data entry requirements (245 [86.9%]), long cut-and-pasted notes (212 [75.2%]), inaccessibility of information from multiple institutions (206 [73.1%]), notes geared toward billing (206 [73.1%]), interference with work-life balance (178 [63.1%]), and problems with posture (144 [51.1%]) and pain (134 [47.5%]) attributed to the use of EHRs. Overall, EHR design and use factors accounted for 12.5% of variance in measures of stress and 6.8% of variance in measures of burnout. Work conditions, including EHR use and design factors, accounted for 58.1% of variance in stress; key work conditions were office atmospheres (β̂ = 1.26; P < .001), control of workload (for optimal control: β̂ = −7.86; P < .001), and physical symptoms attributed to EHR use (β̂ = 1.29; P < .001). Work conditions accounted for 36.2% of variance in burnout, where challenges included chaos (adjusted odds ratio, 1.39; 95% CI, 1.10-1.75; P = .006) and physical symptoms perceived to be from EHR use (adjusted odds ratio, 2.01; 95% CI, 1.48-2.74; P < .001). Coping strategies were associated with only 2.4% of the variability in stress and 1.7% of the variability in burnout.
Conclusions and Relevance
Although EHR design and use factors are associated with clinician stress and burnout, other challenges, such as chaotic clinic atmospheres and workload control, explain considerably more of the variance in these adverse clinician outcomes.
The adoption of the electronic health record (EHR) has occurred alongside the dramatic and troubling rise in clinician stress and burnout.1-3 This association has fueled the debate over the extent to which EHRs are associated with the epidemic of clinician stress and burnout. Technostress (ie, the stress related to technological tools in numerous industries) is real,4 but the degree to which it is a factor in medicine is largely unknown.
The introduction of EHRs has resulted in shifting many clerical tasks to clinicians (eg, billing, coding, and quality control) as well as creating new tasks to be performed during clinical encounters (eg, data entry, computerized decision support, computerized order entry, and electronic prescribing). These new tasks have increased the cognitive and physical load on the clinician in many ways.5,6 For example, e-prescribing, which has benefits, has also created an additional burden by requiring clinicians to know where to route prescriptions at the time they prescribe. This may be a relatively small burden, but repeated multiple times per day and added to the myriad other tasks shifted to clinicians, these technology-enabled tasks have considerably increased clinician workload. In fact, an entirely new medical scribe industry has arisen in order to ameliorate the additional workload.7
We designed this study (Minimizing Stress, Maximizing Success of the Electronic Health Record) to identify the relative contribution of aggregated EHR burdens compared with other burdens (ie, workplace chaos, control of workload) associated with clinician stress and burnout. This work is based on a conceptual framework derived from prior work (Figure).8 Our hypothesis was that EHR-associated stress adds to overall stress and could lead to burnout—which may play a role in the quality of patient care. In this study, we aim to understand which EHR design and use factors are associated with stress and burnout. The potentially challenging EHR design and use factors included in the survey instrument were identified through physician focus groups conducted in the first phase of the study.9 The design and use factors studied were intentionally limited to those over which clinicians and their institutions might have some control. This in no way minimizes other societal factors, such as governmental regulation and malpractice, that could be associated with clinician stress and burnout.10-12 This survey phase of our study quantifies the association of these EHR design and use factors with clinician stress and burnout to address the following questions: (1) what specific EHR design and use factors are most strongly associated with clinician stress and burnout? (2) What amount of overall stress and burnout is associated with EHRs? And, (3) what coping strategies or organizational solutions did respondents feel are important in addressing stress and burnout?
The methods for this study have been previously reported.9 In brief, physician focus groups at 3 institutions (Stanford Hospital and Clinics, Stanford, California; University of New Mexico, Albuquerque; and Centura Health Physician Group, Westminster, Colorado) identified EHR design and use factors that were perceived as successful and those that were associated with user stress, burnout, or unintended physical symptoms. We also identified commonly used coping strategies by the clinicians.
The EHR design and use factors identified in prior clinician focus groups informed the design of the survey instrument, which is freely available.13 The instrument included questions from previously validated instruments to measure stress, burnout, and other challenges identified by Motowidlo,14 the Physician Worklife Survey,15 the Minimizing Error, Maximizing Outcome Study,16 and the Healthy Work Place Study.17,18 Questions also focused on workplace characteristics such as workload control19 and work atmosphere (a single item measure from the Minimizing Error, Maximizing Outcome Study)20 as well as patient complexity and organizational culture, including value alignment between leaders and clinicians. This survey study complied with the American Association for Public Opinion Research (AAPOR) reporting guideline.
The study survey instrument was pilot tested on 10 clinicians at Hennepin County Medical Center (Minneapolis, Minnesota). We then deployed the finalized instrument in 2 waves at the 3 focus group sites from August 9, 2016, through July 7, 2017. The institutional review boards at all participating institutions approved the study, and completing the survey was considered providing consent.
We used REDCap version 8.10.7 (Vanderbilt University) to deploy an electronic version of the instrument. Nonresponders to the REDCap electronic survey were mailed paper instruments. The electronic instrument used continuous slider bars for respondents to indicate a score from 0 to 100, where 0 indicated not at all and 100 indicated to a great extent. The paper instrument used Likert scales mapped to the scale of 0 to 100 for analysis (ie, 1, not at all, mapped to 15; 2 mapped to 40; 3 mapped to 60; and 4, to a great extent, mapped to 85).
The survey’s design attempted to determine the following: (1) perceived EHR successes, (2) EHR design and use factors associated with clinician stress and burnout, (3) perceived adverse personal outcomes (eg pain or anxiety), (4) things that could improve the EHR experience (eg, greater staff support, scribes, or fewer clicks per task), and (5) coping strategies (eg, exercise or setting boundaries). We sampled clinicians (physicians and advanced practice clinicians, including nurse practitioners and physician assistants) at 3 institutions from 5 disciplines: general internal medicine, medical subspecialties, general pediatrics, pediatric subspecialties, and family medicine. We excluded residents, as we thought they could have dissimilar experiences of stress and burnout than practicing clinicians. We determined respondent stress levels using the 4-item validated measures from Motowidlo,14 a continuous measure that ranges from 4 to 20, and burnout using the single-item validated measure from the Physician Worklife Study, in which a score of 3 or more indicates burnout.21 While a binary approach to burnout has been controversial,22,23 this measure has been used and validated in many settings and among thousands of respondents for 20 years, and it is associated with adverse work conditions and adverse clinician outcomes, such as intent to leave the practice. We ran additional analyses using the 5-choice measure of burnout as an ordered categorical (as opposed to binary) outcome and found no substantive differences between the 2 methods.
Answers to survey questions were analyzed as standard summary statistics. We reported continuous variables as mean and SD and categorical variables as number of respondents and percentages of total sample.
Linear regression was used to determine the association of focus group–identified variables (eg, work conditions, EHR design and use factors, and coping strategies) with clinician-reported stress, which we scored according to the Motowidlo 4-item measure,14 and burnout. β̂ was used to estimate the magnitude and direction of association, and it was calculated using the least-square estimation technique. We used logistic regression with stepwise selection, which is a combination of the forward and backward selection techniques, to estimate the association of focus group–identified variables with the odds of clinician-reported burnout, which we measured as a binary outcome based on a single question (with burnout representing endorsement of any choice with the word burnout in it).14 We used construct variables created to summarize the associations of variables within the same domain with stress and burnout. To develop the final regression model for stress, variables with R2 greater than 0.10 in the univariate analysis or that were determined to be of special interest were considered candidate variables for the multivariable model. The final logistic regression model for burnout used a stepwise selection technique, which was determined to be the most comprehensive method because it combines both forward and backward selection. To justify lumping together different types of clinicians and specialties, 1-way analysis of variance was used to examine if statistically significant differences existed in the means of outcome measures across clinician type (ie, MD, DO, nurse practitioner, or physician assistant) or specialty (ie, primary care, nonprocedural specialist, or procedural specialist). Diagnostics done on the regression and logistic models were the Breusch-Pagan test for constant variance and the Hosmer-Lemeshow test for goodness of fit, noting that P > .05 indicates having constant variance for the regression model and correct fit for the logistic model respectively. (These showed that the models were well calibrated.) Finally, we performed a statistical factor analysis using the varimax rotation method on 9 EHR design and use items to summarize the association of EHRs with stress and burnout. We used SAS version 9.4 (SAS Institute, Inc) for all analyses. Statistical significance was set at P < .05, and all tests were 2-tailed.
Between August 2016 and July 2017, we surveyed 640 clinicians from 3 institutions, with 282 (44.1%) responding (208 [73.8%] electronically and 74 [26.2%] on paper); 160 (56.7%) were women, 241 (85.5%) were physicians (MDs and DOs), and 193 (68.4%) worked in primary care (Table 1). Overall, 256 respondents (90.8%) answered at least 95 of the 105 survey questions. The 1-way analysis of variance showed no significant difference in mean (SD) burnout between clinician types (DO, 2.33 [0.52]; MD, 2.54 [0.94]; nurse practitioner, 2.14 [0.53]; physician assistant, 2.45 [0.94]; P = .42) or between practice types (primary care, 2.51 [0.52]; nonprocedural specialist, 2.48 [0.82]; procedural specialist 2.59 [0.76]; P = .86). Therefore, neither of these components was controlled for in the analysis. Most participants noted stressful work conditions: 210 (74.5%) reported time pressure for documentation, and 170 (60.2%) spent moderately high or excessive time on the EHR at home (Table 1). Overall, 142 (50.4%) felt they had insufficient personal time, and 134 (47.5%) reported having minimal coverage for their EHR inboxes when needed. Only 95 (33.7%) reported that their practices emphasized work-life balance, while 215 (76.2%) said that productivity was overemphasized. Half (140 [49.6%]) reported marginal or poor control over workload, and 143 (50.7%) judged their office atmospheres as chaotic or tending toward chaotic. Almost half (127 [45.0%]) described symptoms of burnout, and 117 (41.5%) indicated they were moderately to definitely likely to leave their practices within 2 years (Table 1).
The EHR successes participants identified included the ability to message colleagues electronically (197 [69.9%]), access to the EHR from home (213 [75.5%]), and the opportunity to share results with patients (136 [48.2%]). The most troublesome EHR design and use factors reported were excessive data entry requirements (245 [86.9%]), “note bloat” (unnecessarily long cut-and-pasted progress notes; 212 [75.2%]), inaccessible information from other institutions (206 [73.1%]), notes geared toward billing rather than patient care (206 [73.1%]), problems with work-life balance (178 [63.1%]), and 2 physical items that respondents attributed to EHR use: posture issues (144 [51.1%]) and pain (134 [47.5%]).
The EHR design and use factors significantly associated with high clinician stress were information overload (β̂ = 0.37; P < .001), slow system response times (β̂ = 0.42; P < .001), excessive data entry (β̂ = 0.43; P < .001), inability to navigate the system quickly (β̂ = 0.38; P < .001), note bloat (β̂ = 0.24; P = .01), fear of missing something (β̂ = 0.34; P < .001), interference with the patient-clinician relationship (β̂ = 0.29; P < .01), and notes geared toward billing (β̂ = 0.41; P < .001) (Table 2). In our analyses, burnout was used as a dichotomous as well as an ordered categorical variable, and there were no substantive differences between the 2 approaches. All of the previously listed EHR design and use factors were independently associated with burnout except fear of missing something. These factors collectively accounted for 12.5% and 6.8% of the variance in stress and burnout (as a binary outcome), respectively. Physical symptoms attributed to EHR use increased odds of burnout (adjusted odds ratio [aOR], 2.01; 95% CI, 1.48-2.75; P < .001)
Factors not related to EHRs associated with high levels of variance in stress were office atmospheres (β̂ = 1.26; P < .001), control of workload (for optimal control: β̂ = −7.86; P < .001), time for personal and family life (for disagree: β̂ = −2.30; P < .001), time for documentation at work (for satisfactory: β̂ = −2.93; P < .001), value alignment with leaders (for agree strongly: β̂ = −4.73; P < .001), professional and personal life balance (β̂ = −1.56; P < .001), physical symptoms attributed to EHR use (β̂ = 1.29; P < .001) and hours worked per week (β̂ = 0.78; P < .001). Within a multivariable linear regression model (Table 3), these variables, along with the EHR design and use factors listed in Table 2, consequences of EHR use, and EHR use at home, accounted for 58.1% of variance in clinician-reported stress and 36.2% of variance in burnout (Table 4). A chaotic work environment increased the odds of burnout (aOR, 1.39; 95% CI, 1.10-1.75; P = .006).
Coping strategies for reducing stress felt to be associated with the EHR included talking with others (194 [68.8%]), exercise (192 [68.1%]), setting work boundaries (161 [57.1%]), discussing EHR messages with others rather than pinging electronic messages back and forth (149 [52.8%]), and writing shorter notes (142 [50.4%]). As a combined variable, coping strategies accounted for only 2.4% and 1.7% of the variability in stress and burnout respectively (data not shown). Setting boundaries (β̂ = −0.02; P < .01) and taking breaks (β̂ = −0.02, P = .006) were independently associated with reductions in overall stress, while exercise (aOR, 0.99; 95% CI, 0.98-1.00; P = .04) and taking breaks (aOR, 0.99; 95% CI, 0.98-1.00; P = .003) were associated with reductions in the odds of burnout.
We performed a statistical factor analysis using the varimax rotation method on the 9 EHR design and use factors listed in Table 2. We found that the first 2 statistical factors from the factor analysis accounted for 52.2% of the variability in EHR design and use items. We characterize these 2 factors as follows: (1) interference with patient care (eg, note bloat, interference with patient-clinician relationships, and notes geared toward billing) and (2) inefficient systems (eg, slow system response times, inability to navigate the system quickly, and excessive data entry). Thus, more than half of the variance in EHR issues associated with clinician stress and burnout stemmed from interference with patient care and inefficient EHR systems.
In this cross-sectional survey of 282 clinicians from 3 health systems, we identified 7 EHR design and use factors associated with high stress and burnout. These were information overload, slow system response times, excessive data entry, inability to navigate the system quickly, note bloat, interference with the patient-clinician relationship, fear of missing something, and notes geared toward billing. While previous studies have identified several of these EHR design and use items as challenging to clinicians,9,24,25 we believe this study is the first to show an association between these factors and objectively validated stress and burnout scales.
In this study, 45.0% of participants described symptoms of burnout, consistent with the findings of the national survey by Shanafelt et al2 in which 44% of physicians reported at least 1 symptom of burnout. The amounts of variation in stress and burnout associated with the EHR design and use factors listed in Table 2 were 12.5% and 6.8%, respectively. Thus, other sources of burnout aside from the EHR (such as lack of control of workload, chaotic environments, lack of attention to work-life balance, and ineffective teamwork) will also need to be addressed as medical practices seek to reduce burnout.
Many of the identified EHR design and use factors may be remediable through a combination of improvements by EHR vendors, local improvements by information technology personnel, and training of clinicians in the clinical environment. However, some of the identified factors may require higher-level actions on the part of clinic or governmental policy makers, for example, by allowing notes to be more geared toward clinical care than billing practices. Documentation requirements for billing purposes is an EHR design characteristic associated with both stress and burnout. The length of clinical notes has essentially doubled since the enactment of the Health Information Technology for Economic and Clinical Health Act.26 Physicians outside the United States are more likely to report satisfaction with their EHRs, where clinical documentation is significantly shorter and contains much less information in support of billing and compliance.26 The American Medical Informatics Association has recently called for a long-term strategy from the US Department of Health and Human Services to decouple clinical documentation from billing, regulatory, and administrative compliance requirements.27
Information overload may be associated with EHR design in which too much clinically unnecessary information is displayed. The aviation industry has a user interface design philosophy called quiet dark, where information is not displayed until something goes wrong or needs the pilot’s attention.28 In other words, the default state of all indicator lights is off during normal conditions. Applying this philosophy to EHR design could potentially reduce the amount of unnecessary data displayed based on particular users’ need and context, reducing the information overload problem. Arguably, the current state of EHR design is loud bright, where virtually all information, normal or otherwise, appears in relatively the same manner regardless of its importance to the clinician or patient. Although abnormal results from laboratory tests are highlighted, all normal values are typically displayed and occupy the same amount of space and are given the same prominence as abnormal results. Given the proliferation of standardized templates as a time-saving tool for data entry, the amount of unnecessary, repetitive, normal information (ie, note bloat) is increasing vs a design where an economy of information relevant to the patient’s current needs and context is used.29
The data entry problem has created the scribe movement and produced promising results, at least in terms of clinician and patient satisfaction.30 However, scribes only help with data entry during office visits and not with EHR tasks at other times and in other venues. A more comprehensive approach is to use specially trained medical assistants (MAs) to relieve the clinician from clinic tasks (eg, responding to routine in-basket messages, refilling some prescriptions per protocol, completing paperwork). Before the clinician meets the patient, the MA completes prework (eg, medication reconciliation, review of systems, documentation of chief concern, and any protocolized clinical measurements, such as peak flows or pulse oximetry). The MA scribes during the clinician encounter, and after the clinician leaves the room, the MA can review the plan of care, deliver patient education, process referral requests, and schedule follow-up appointments.31
Some of the troublesome EHR design and use factors, such as the inability to navigate the system quickly, are attributable to computer-human interaction problems. In fact, most of the current EHR user interface designs are still based on 2-dimensional paper metaphors (eg, tabs, flowsheets, tables, and forms) and do not take advantage of the potential of graphics capabilities now in the most basic computers.32 More research to determine what display metaphors beyond paper are most efficient could help. Complaints of interference with the patient-clinician relationship is evidence that clinicians are troubled by their excessive focus on the screen rather than the patient. While most studies have shown the presence of the EHR in the exam room does not adversely affect patient satisfaction,9,33,34 clinicians feel that EHRs requiring clinically irrelevant data entry take away from their relationships with their patients.35 Our study shows that this is significantly associated with clinician stress and burnout.
The proportion of clinicians reporting pain (47.5%) and posture issues (51.1%) attributed to EHR use was high. Ergonomics are rarely addressed in most clinical settings. Clinicians often must work at several workstations, with different heights and seat structures. Collaboration with employee health groups skilled at ergonomics could potentially have a substantive effect on the health outcomes of our clinician workforce.36 This is an area ripe for further quality improvement studies.
Coping strategies clinicians suggested to reduce EHR-associated stress included exercise (used by 68.1% of our sample), verbally discussing issues with other clinicians (68.8%), and setting boundaries for work while at home (57.1%). Setting boundaries, exercise, and taking breaks were significantly associated with reductions in overall stress and burnout and may be useful components to incorporate into stress reduction interventions. It is not clear how many of these strategies clinicians actually used or how effective they were at using them.
The strengths of this study include surveying a diverse group of clinicians, including academic, community-based, and rural institutions and practices, physicians and advanced practice clinicians, and a mix of specialists and nonspecialty ambulatory care clinicians. In addition, the list of the EHR design and use factors the clinicians rated in the survey was defined by clinicians in multi-institutional focus groups.9 The survey response rate (44.1%) was reasonable for large clinician-based studies with no financial incentive. The design of the instrument included questions previously validated in studies of physicians about stress and burnout.
This study has limitations, including its cross-sectional nature and the use of self-reported metrics. One needs to consider response bias, given the 44.1% response rate. The relatively modest sample size limits validity. As respondents came from only 3 institutions, these results may not be more widely generalizable. The mapping of the paper instrument’s Likert scales to the REDCap slider bars scale may have introduced some bias. Despite using validated instruments to measure burnout and stress, the survey relied on the respondents’ own definitions. Self-reported metrics may underrepresent the numbers at risk. As Knox et al37 found, a self-defined, single-item burnout measure identified significantly fewer physicians most at risk of burning out compared with the Maslach Burnout Inventory. All respondents were grouped together for this analysis, which does not account for possible intragroup differences, such as between physicians and advanced practice clinicians.37
Stress and burnout associated with EHRs is prevalent and may be at least partly remediable at the local level. The issues identified in our list of EHR-associated challenges may provide designers, government regulators, and clinical leaders with targets for improvement of EHR design. Other work conditions are associated with stress and burnout in clinicians and deserve equal attention.
Accepted for Publication: July 1, 2019.
Published: August 16, 2019. doi:10.1001/jamanetworkopen.2019.9609
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2019 Kroth PJ et al. JAMA Network Open.
Corresponding Author: Philip J. Kroth, MD, MS, Health Sciences Library and Informatics Center, University of New Mexico, MSC09 5100, One University of New Mexico, Albuquerque, NM 87131-0001 (firstname.lastname@example.org).
Author Contributions: Drs Kroth and Linzer had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Kroth, Morioka-Douglas, Veres, Linzer.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Kroth, Poplau, Parshall, Corrigan, Linzer.
Critical revision of the manuscript for important intellectual content: Kroth, Morioka-Douglas, Veres, Babbott, Qeadan, Linzer.
Statistical analysis: Qeadan, Parshall.
Obtained funding: Kroth, Veres, Poplau, Linzer.
Administrative, technical, or material support: Kroth, Veres, Corrigan, Linzer.
Conflict of Interest Disclosures: Dr Linzer reported receiving support for wellness research, training wellness champions, and scientific oversight from the American Medical Association, the American College of Physicians, the Institute for Healthcare Improvement, and the Association of Chiefs and Leaders in General Internal Medicine; receiving funds paid for salary, which went to Hennepin County Medical Center; and honoraria from Brown University and the University of Chicago, which were donated to the Health Foundation at Hennepin County Medical Center. No other disclosures were reported.
Funding/Support: This project was supported by grant number R18HS022065 from the Agency for Healthcare Research and Quality.
Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the Agency for Healthcare Research and Quality.
Additional Contributions: Gale Hannigan, PhD (New Mexico Health Sciences Library and Informatics Center, Albuquerque, New Mexico), assisted with the multiple literature searches and text editing of this work, and Jeremiah Menk, MS (University of Minnesota, Minneapolis), assisted with biostatistics in the early phases of this project. They were compensated for their time.
Create a personal account or sign in to: