Advertisement
Mayo Clinic Proceedings Home

Optimization Sprints: Improving Clinician Satisfaction and Teamwork by Rapidly Reducing Electronic Health Record Burden

Open AccessPublished:February 26, 2019DOI:https://doi.org/10.1016/j.mayocp.2018.08.036

      Abstract

      Objective

      To evaluate a novel clinic-focused Sprint process (an intensive team-based intervention) to optimize electronic health record (EHR) efficiency.

      Methods

      An 11-member team including 1 project manager, 1 physician informaticist, 1 nurse informaticist, 4 EHR analysts, and 4 trainers worked in conjunction with clinic leaders to conduct on-site EHR and workflow optimization for 2 weeks. The Sprint intervention included clinician and staff EHR training, building specialty-specific EHR tools, and redesigning teamwork. We used Agile project management principles to prioritize and track optimization requests. We surveyed clinicians about EHR burden, satisfaction with EHR, teamwork, and burnout 60 days before and 2 weeks after Sprint. We describe the curriculum, pre-Sprint planning, survey instruments, daily schedule, and strategies for clinician engagement.

      Results

      We report the results of Sprint in 6 clinics. With the use of the Net Promoter Score, clinician satisfaction with the EHR increased from −15 to +12 (−100 [worst] to +100 [best]). The Net Promoter Score for Sprint was +52. Perceptions of “We provide excellent care with the EHR,” “Our clinic’s use of the EHR has improved,” and “Time spent charting” all improved. We report clinician satisfaction with specific Sprint activities. The percentage of clinicians endorsing burnout was 39% (47/119) before and 34% (37/107) after the intervention. Response rates to the survey questions were 47% (97/205) to 61% (89/145).

      Conclusion

      The EHR optimization Sprint is highly recommended by clinicians and improves teamwork and satisfaction with the EHR. Key members of the Sprint team as well as effective local clinic leaders are crucial to success.

      Abbreviations and Acronyms:

      APP (advanced practice provider), EHR (electronic health record), NPS (Net Promoter Score), PEP (physician efficiency profile)
      Electronic health record (EHR) implementation rates have increased
      • DesRoches C.M.
      • Charles D.
      • Furukawa M.F.
      • et al.
      Adoption of electronic health records grows rapidly, but fewer than half of US hospitals had at least a basic system in 2012.
      • Adler-Milstein J.
      • DesRoches C.M.
      • Furukawa M.F.
      • et al.
      More than half of US hospitals have at least a basic EHR, but stage 2 criteria remain challenging for most.
      and so has clinician EHR burden, which has been described as the computer-based clerical work associated with patient care.
      • Arndt B.G.
      • Beasley J.W.
      • Watkinson M.D.
      • et al.
      Tethered to the EHR: primary care physician workload assessment using EHR event log data and time-motion observations.
      Clinician EHR burden has been cited as a major cause of clinician burnout.
      • Shanafelt T.D.
      • Hasan O.
      • Dyrbye L.N.
      • et al.
      Changes in burnout and satisfaction with work-life balance in physicians and the general US working population between 2011 and 2014.
      • Shanafelt T.D.
      • Dyrbye L.N.
      • Sinsky C.A.
      • et al.
      Relationship between clerical burden and characteristics of the electronic environment with physician burnout and professional satisfaction.
      • Friedberg M.W.
      • Chen P.G.
      • Van Busum K.R.
      • et al.
      Factors affecting physician professional satisfaction and their implications for patient care, health systems, and health policy.
      Burnout is characterized by a loss of enthusiasm for work, feelings of cynicism, and a low sense of personal accomplishment.
      • Bodenheimer T.
      • Sinsky C.
      From triple to quadruple aim: care of the patient requires care of the provider.
      In addition to its toll on individuals and their families, clinician burnout may negatively affect institutional finances. Clinician burnout can lead to early retirement. It is estimated that when a physician retires early, $250,000 of productivity is lost per year
      • Dewa C.S.
      • Jacobs P.
      • Thanh N.X.
      • Loong D.
      An estimate of the cost of burnout on early retirement and reduction in clinical hours of practicing physicians in Canada.
      and that it may cost $500,000 to $1 million
      • Noseworthy J.
      • Madara J.
      • Cosgrove D.
      • et al.
      Health Affairs Blog. Physician burnout is a public health crisis: a message to our fellow health care CEOs.
      to replace a physician. An organization with 500 physicians could potentially spend more than $6 million annually replacing burned out physicians.
      • Shanafelt T.
      • Goh J.
      • Sinsky C.
      The business case for investing in physician well-being.
      To ease the clinician EHR burden, organizations have tried to boost clinician efficiency in using the EHR
      • Bohman B.
      • Dyrbye L.
      • Sinsky
      • et al.
      NEJM Catalyst. Physician well-being: the reciprocity of practice efficiency, culture of wellness, and personal resilience.
      by disseminating EHR tip sheets and holding various types of training sessions. At our institution, similar tactics have generally failed to increase either clinician efficiency or satisfaction with the EHR. Previously, given limited resources, we had prioritized EHR improvements that affected the largest number of clinicians. As a result, interventions that clinicians highly desired, such as specialty-specific improvements in the EHR and individualized EHR training, were not approved or funded, leading to further clinician dissatisfaction. Consequently, we recently designed, implemented, and assessed a novel intervention, the optimization “Sprint,” an intensive team-based intervention designed to reduce clinician EHR burden, alleviate clinician burnout, and improve clinician satisfaction with the EHR, one clinic at a time.

      Methods

      Please see the Supplemental Material (available online at http://www.mayoclinicproceedings.org) for an expanded Methods section that includes meeting and huddle agendas, workflow redesign tactics, clinic-specific and system-wide decision-making strategies, a detailed grid of daily Sprint team activities, and pre- and post-Sprint questionnaires.

      Setting

      UCHealth is a large integrated health network in Colorado, consisting of 400 clinics and 9 hospitals, all using a single EHR (version 2015, Epic Systems). Sprints were carried out in 6 clinics (Table 1). The Colorado Multiple Institutional Review Board deemed the activities of this project to be quality improvement and exempt from full review by the institutional review board.
      Table 1Specialties of Clinicians and Clinical Staff
      AMC = Academic Medical Center; APP = advanced practice provider; DO = osteopath; MA = medical assistant; MD = physician; MG = medical group (community); NP = nurse practitioner; PA = physician assistant; RN = registered nurse.
      ClinicNo. of clinicians (MD, DO, NP, PA, and midwife)No. of clinical staff (RN, MA, and clerk)
      AMC endocrine2313
      MG neurology clinic 154
      AMC neurology9628
      MG hematology87
      MG neurology clinic 2117
      AMC obstetrics and gynecology90 (−13)
      13 clinicians who received surveys did not participate in Sprint as their oncology subspecialty was later excluded.
      38
      Total physicians167
      Total APP (NP, PA, and midwife)66
      Total233 (220)
      233 clinicians received pre- and post-surveys, and 220 clinicians actually participated in Sprint.
      97
      a AMC = Academic Medical Center; APP = advanced practice provider; DO = osteopath; MA = medical assistant; MD = physician; MG = medical group (community); NP = nurse practitioner; PA = physician assistant; RN = registered nurse.
      b 13 clinicians who received surveys did not participate in Sprint as their oncology subspecialty was later excluded.
      c 233 clinicians received pre- and post-surveys, and 220 clinicians actually participated in Sprint.

      Intervention

      The Sprints intervention had 3 primary components: (1) training clinicians to use existing EHR features more efficiently, (2) redesigning the multidisciplinary workflow within the clinic, and (3) building new specialty-specific EHR tools. Sprint was a quality improvement intervention, and we prioritized continuous process improvement over consistent data collection. Thus, we adjusted some survey questions over time to provide better insight, and thus the total number of respondents to some questions will differ from others. For each of the 6 clinics, the most updated survey at the beginning of that Sprint was sent to all clinicians in that clinic both before and after Sprint.

      Sprint Timeline

      Sprints were carried out between January 16, 2016, and July 21, 2017. The Sprint leaders met with clinic leaders 90, 60, and 30 days before each Sprint to prepare for the intervention, identify EHR frustrations, and prioritize potential solutions. For pre-Sprint meetings and Sprint on-site details, see the Supplemental Material (available online at http://www.mayoclinicproceedings.org). During the course of each Sprint, all clinicians had their time protected so that they could participate in a 2-hour kickoff meeting, three 1-to-1 training sessions, and a 2-hour wrap-up session.

      Agile

      We used Agile methodology
      Scrum.org
      The Home of Scrum: An Advanced Scrum Master Class.
      as a guiding strategy for Sprint. Agile methodology improves software delivery by focusing on the voice of the customer and making rapid incremental improvements in a short time period. We held daily huddles and prioritized tasks on the basis of feedback from clinicians and clinic staff. We used scrum boards to depict progress in fulfilling requested EHR changes. When possible, we facilitated conversations between specialists in the clinic undergoing Sprint and specialists at similar clinics in our organization to build consensus on specialty-specific EHR tools.

      Training Content

      The training content evolved over several iterations of Sprints. The physician informaticist leader built 10 sets of efficiency tips (the last 10 activities in Table 2).
      Table 2Clinician Assessment of Helpfulness of Sprint Activities
      EHR = electronic health record; PI = physician informaticist.
      ActivityDescription/contentPercentage of respondents who responded positively (No. of respondents)
      The percentage of clinicians who responded “agree” or “strongly agree” to the statement that the listed activity was helpful. A total of 186 clinicians were surveyed, and 84 responded to the survey. Not all clinicians participated in each activity. Total respondents to each question are shown within parentheses.
      1-to-1 trainingClinician met individually with a trainer or PI to learn specific skills or address personal frustrations with the her93% (68)
      Speech recognition toolClinician learned how to use speech recognition with the EHR to improve word accuracy and create navigation shortcuts87% (54)
      New or redesigned toolsSpecialty-specific EHR tools built during Sprint in response to clinic requests80% (66)
      Notes: smart phrasesCreation of personalized note templates to autotype frequently used phrases and allow efficient selection from drop-down lists80% (64)
      Observation/shadowTrainer or PI observed a clinician use the EHR in an examination room or work area and offered feedback79% (58)
      Chart review efficiencyTools for finding patient information in the EHR, including “chart search” and “custom filters”77% (62)
      Notes: problem listProblem-based charting, problem list sorting and maintenance, and autocorrect dictionary73% (67)
      In-basket: clinic messagingManaging patient calls, prescription renewals, and communication with referring physicians and receiving faxes and other paper forms66% (59)
      Ordering efficiencyMaximizing efficiency in placing single orders, multiple orders, future orders, and favorite orders65% (58)
      Out-of-office workflowsBest practices for notifying patients and for EHR in-basket (messaging) coverage when the clinician is not available62% (48)
      In-basket: test resultsManaging test results from internal and external sources and notifying patients60% (60)
      Medication managementEfficiency tips for prescribing, setting preferences, managing refills, adjusting doses of existing medications, and reconciling medications56% (59)
      Check-in, check-out workflowsCoordinating care with clinic staff at patient check-in (verifying referring physician and preferred pharmacy) and check-out (follow-up, referrals, and testing)48% (48)
      a EHR = electronic health record; PI = physician informaticist.
      b The percentage of clinicians who responded “agree” or “strongly agree” to the statement that the listed activity was helpful. A total of 186 clinicians were surveyed, and 84 responded to the survey. Not all clinicians participated in each activity. Total respondents to each question are shown within parentheses.

      Multidisciplinary Workflow Redesign

      Some Sprint team members met individually and in small groups with nonclinician staff to observe the multidisciplinary workflow and patient flow in the clinic. Using this information, trainers taught EHR best practices to staff, and the clinic manager redesigned common workflows so that all clinicians and staff did things the same way. For examples of redesigned workflows, see the Discussion section.

      Building New EHR Tools

      Once EHR tool requests were approved by clinic leaders, the Sprint project manager placed them in a visual chart under the following headings: Backlog, To Do, In Process, Done, and Parking Lot. These items were reprioritized and repositioned daily. In this way, clinicians and clinic staff could see the status of all requests, including Parking Lot items that would not be addressed during Sprint.

      Fixing Existing EHR Tools

      Broken items that were fixed by EHR analysts included flow sheets that no longer calculated correctly, smart links to test results that did not display requested data, and incorrect routing of patient portal messages. The Sprint team also facilitated the repair or replacement of EHR equipment such as keyboard trays, printers, monitors, and radiofrequency identification badge readers for EHR login.

      Personnel

      The Sprint Team

      The Sprint team consisted of 1 full-time equivalent primary physician informaticist (a role shared by 2 physicians), a nurse informaticist, a project manager, 4 EHR trainers (nonclinician staff certified to teach EHR tools), and 4 EHR analysts. The physician informaticist helped translate clinician requests for EHR improvements into specific technical changes. Having the physician informaticist translate requests reduced misunderstandings and eliminated many hours of wasted EHR analyst effort. For example, a clinician’s request for a complex cardiac monitoring order took up dozens of hours of EHR analyst time. Despite this, the requestor was still unsatisfied. After a physician informaticist became involved, a new design strategy was implemented, which satisfied both the requestor and the analyst. By involving the physician informaticist earlier in the request and design process, we estimated that dozens of hours were saved in the process of handling subsequent requests. The nurse informaticist helped redesign multidisciplinary workflows and trained clinic staff. The Sprint project manager coordinated pre-Sprint meetings, was present in the clinic for the entire Sprint, and coordinated the activities of the 4 clinics simultaneously involved in various stages of the Sprint process.

      The Clinical Operations Team

      Before Sprint, organizational operational leaders (the vice president of ambulatory services and the ambulatory clinic directors) worked with the Sprint leaders to select clinics for Sprint according to each clinic’s desire for EHR improvement and the presence of effective clinic leadership. In larger academic specialty clinics, each subspecialty designated one “clinical content lead” clinician to build consensus within their group, prioritize requests, communicate with the Sprint team, and review newly built EHR tools.

      Assessment of Clinician Participation, Satisfaction, and Burnout

      Participation

      The Sprint team tracked individual clinician attendance at all Sprint activities. A large grid that displayed activities and clinician names made it easy to identify clinicians who were not participating and give them additional attention.

      Satisfaction With Sprint and the EHR

      We sent all clinicians a pre-Sprint survey created with SurveyMonkey.com 60 days before their clinic’s Sprint and a post-Sprint survey 2 weeks after the Sprint process concluded. Survey responses were confidential and anonymous For survey instruments, see the Supplemental Material (available online at http://www.mayoclinicproceedings.org). Clinic leaders reminded clinicians to complete the surveys 3 times over a 2-week period. Clinician satisfaction with the Sprints intervention and their satisfaction with the EHR before and after Sprint were assessed by calculating the Net Promoter Score (NPS).
      • Reichheld F.F.
      The one number you need to grow.
      Qualtrics
      What is Net Promoter Score (NPS)?.
      Insights From Analytics. Top 10 U.S. Net Promoter Scores (NPS) for 2013.
      Net Promoter System®
      Companies that use net promoter.
      The NPS could range from −100 (worst) to +100 (best).
      In addition, in the survey we asked clinicians to assess their EHR clinical processes before and after Sprint. We also reviewed “physician efficiency profile (PEP)” EHR metrics that measured actual activity in the EHR. Lastly, all 186 clinicians in the academic neurology and obstetrics and gynecology clinics were asked to rate on a Likert-like scale the value of specific Sprint training activities.

      Burnout

      In both the pre- and post-Sprint surveys, we presented clinicians with the “emotional exhaustion domain” item from the Maslach Burnout Inventory.
      • West C.P.
      • Dyrbye L.N.
      • Sloan J.A.
      • Shanafelt T.D.
      Single item measures of emotional exhaustion and depersonalization are useful for assessing burnout in medical professionals.
      We considered the emotional exhaustion of burnout to be present if the response was at least once a week in frequency. For the remainder of this article, we use “burnout” to represent the emotional exhaustion domain from the Maslach Burnout Inventory.
      • West C.P.
      • Dyrbye L.N.
      • Sloan J.A.
      • Shanafelt T.D.
      Single item measures of emotional exhaustion and depersonalization are useful for assessing burnout in medical professionals.

      Narrative Comments

      The e-mail surveys included opportunities for clinicians to provide narrative feedback. Although in the clinics, Sprint team members also captured clinician comments about the Sprint process and its effect on their work and home lives.

      Results

      Participation in Sprint

      A total of 220 clinicians participated in Sprint, of whom 143 (65%) attended at least 3 training sessions and 42 (19%) participated in more than 10 sessions. Seventy-seven clinicians (35%) attended fewer than 3 training sessions and 44 (20%) participated in no sessions. Pre- and post-Sprint surveys were sent to a total of 233 clinicians (Table 1). Thirteen clinicians from an oncology subspecialty of obstetrics and gynecology were subsequently excluded from Sprint because of physical distance. The percentage of physicians (32%) and advanced practice providers (APPs) (30%) attending fewer than 3 sessions were similar. The percentage of APPs responding to the survey reflected their prevalence in the clinician population: 29% of all clinicians were APPs and 27% of survey respondents were APPs.

      Satisfaction With Sprint

      We were able to survey clinicians in 5 of 6 clinics after Sprint. Because of an oversight, we did not include the NPS for Sprint question for 1 large clinic. Eighty-nine of 145 clinicians (61%) responded to the post-Sprint survey. For all respondents, the NPS for Sprint was +52 (Figure 1).
      Figure thumbnail gr1
      Figure 1Clinician NPSs for Sprint and the EHR. NPS range: −100 (worst) to +100 (best). Response rates: NPS for Sprint: 61% (89 of 145); NPS for the EHR: before Sprint: 54% (125 of 233), 2 weeks after Sprint: 52% (122 of 233). EHR = electronic health record; NPS = Net Promoter Score.

      Evaluation of Sprint Training Activities

      We surveyed all 186 clinicians from 2 academic clinics about the helpfulness of Sprint training activities, and 84 (45%) responded. The highest-rated activities were 1-to-1 training sessions, speech recognition training, new tools training, note-writing smart phrases, and observation/shadow sessions (Table 2).

      Satisfaction With the EHR

      Before Sprint, the NPS for the EHR was −15, indicating low levels of satisfaction, with 54% of clinicians (125 of 233) responding. Two weeks after Sprint, the NPS for the EHR increased to +12, with 52% clinicians (122 of 233) responding (Figure 1).

      Evaluation of EHR Clinical Processes

      Two hundred five clinicians were asked about the EHR clinical process. Clinicians from the first 2 Sprint clinics did not receive these questions because they were redesigned with feedback after their Sprints. The pre- and post-Sprint response rates to the survey questions about EHR-related clinical processes were 52% (107 of 205) and 47% (97 of 205), respectively. In response to the statement “Our clinic has clear policies on how staff and clinicians can best use the EHR together,” the percentage of affirmative responses (defined as either “agree” or “strongly agree”) increased from 21% at baseline to 53% after Sprint (an increase of 32 percentage points). In response to the statement “The EHR helps us provide excellent care,” the percentage of affirmative responses increased from 57% to 77% (an increase of 20 percentage points). In response to the statement “Our clinic’s use of the EHR has improved in the last few months,” the percentage of affirmative responses increased from 23% to 79% (an increase of 56 percentage points). In response to the statement “The amount of time I spend documenting patient care,” 43% of clinicians indicated a decrease (Figure 2). The percentage of clinicians responding affirmatively to these statements before and after Sprint was similar between academic and community practices (data not shown).
      Figure thumbnail gr2
      Figure 2Clinician perceptions of EHR processes in the clinic. Response rates: pre-Sprint: 52% (107 of 205); post-Sprint: 47% (97 of 205). EHR = electronic health record.
      We reviewed the EHR metric data (PEP). The versions of PEP at that time did not track “minutes in system after 7pm” and did not accurately capture data on APPs, and the measurement tool definitions changed during the time line of our interventions. The pre- and post-Sprint PEP data were thus not comparable.

      Effect of Sprint on Burnout

      We changed our burnout question after our first 2 clinics and report results from the latter 4 clinics. Before Sprint, 119 of 205 clinicians (58%) responded to the burnout question; after Sprint, 107 of 205 (52%) responded. Before Sprint, 39% of clinicians reported feeling burned out; after Sprint, the percentage was 34%. On the basis of chi-square analysis, the P value was .434, and this difference does not reach the predetermined α level of significance.

      Most Frequently Requested New EHR Tools

      Over the 20 months of Sprint, we most frequently built synopsis reports, flow sheets, patient-entered questionnaires, customized note templates, and laboratory and radiology result smart links (Table 3). Tools were built, tested, and deployed while the Sprint team was working in the clinic. Requests for new tools were typically fulfilled within 1 to 2 days, which pleasantly surprised many clinicians. One clinician noted, “This is unexpected. It feels like someone in IT actually cares.”
      Table 3Most Frequently Built New EHR Tools
      EHR = electronic health record.
      EHR toolDescription and example
      Graphic reports
      EHR-specific term is “synopsis.”
      A disease- or specialty-specific table and graphic that combines disparate data and displays it over time. Data can include vitals; medication doses; laboratory, radiology, or pathology results; progress notes; and flow sheets. Example: A pain synopsis that exhibits the relationship between symptom scores, blood pressures, urine tox screen results, and narcotic doses and morphine equivalent daily dose
      Flow sheetsA table that displays symptoms, physician examination findings, or laboratory data over time for a single patient. Flow sheet data can be retrieved for quality improvement or research purposes
      Patient-entered questionnairesA questionnaire completed by patients online in the patient portal before their clinic visit. Responses can be imported into flow sheets or into the clinician’s progress note. Example: Blood pressure readings taken at home
      Customized note templatesTemplates that automatically incorporate flow sheets, chart data, and repetitive statements to improve clinician charting efficiency. Example: “Risks and benefits of the proposed procedure were discussed with the patient …”
      Laboratory and radiology smart linksA tool that finds and facilitates documentation of data such as laboratory and radiology results. Example: A link locates the “the most recent hemoglobin A1c” and inserts it in a progress note
      a EHR = electronic health record.
      b EHR-specific term is “synopsis.”

      Narrative Comments

      Narrative feedback on Sprints was almost uniformly positive. One participating clinician wrote, “You have made me a better mother. I get home in time for dinner with my family.” A clinic medical director stated, “A related outcome … has been the improvement in clinicians’ morale and attitudes.” Another clinician noted, “Your extraordinary team spirit is a model for us and for any team.” “This is like a dream,” observed one medical director, “probably will save me 30 minutes a day in charting and placing orders alone.” A clinic manager noted, “The efficiency of clinicians … has improved considerably” and “Most clinicians were at the functional level that they were initially trained at years ago … I cannot say enough good things about this Sprint.” From the 6 clinics, there were 40 survey comments about Sprint and about a dozen verbal comments recorded by Sprint team members. Of all these, all but 2 were positive statements. The negative statements were as follows: “The fact that it took this much time and investment to do a Sprint to retrain providers, nurses, and support staff to use a system we had already been using for >5 years speaks to the fact that it is TOO COMPLICATED!” and “You can’t help me… this is a terrible system, and I don’t have time to talk to you.”

      Discussion

      Sprint was a short-term, intensive, multidisciplinary intervention designed to improve the experience of clinicians and clinic staff using the EHR, one clinic at a time. On the basis of the results of clinician surveys, we concluded that Sprint was a successful intervention. Clinicians gave the Sprints intervention an NPS of +52, which is similar to ratings by customers of Apple’s iPad (+65) and Amazon.com (+69) in 2013.
      Insights From Analytics. Top 10 U.S. Net Promoter Scores (NPS) for 2013.
      The NPS for the EHR improved from −15 at baseline to +12 after Sprint. The emotional exhaustion measure of burnout decreased from 39% to 34% after Sprint, a statistically nonsignificant difference. However, because more than 50% of the entire population responded, a 5% reduction may indeed be a nonrandom improvement. Nevertheless, we theorize that it may take longer than a 2-week practice efficiency intervention to substantially improve clinician burnout.
      • Lyon C.
      • English A.F.
      • Chabot Smith P.
      A team-based model that improves job satisfaction.
      After participating in Sprint, 44% of clinicians indicated that the number of hours spent charting in the EHR had reduced. After Sprint, more clinicians reported that the clinic team was working well together, providing excellent care, and that their use of the EHR had improved. This intervention exhibits that a clinic-based, user-focused intervention to reduce EHR burden and improve clinician satisfaction with the EHR is both feasible and effective.
      The Sprint method may be associated with other positive outcomes, including specialty-specific EHR customization and improved clinical workflow, which can be difficult to achieve with more conventional approaches to EHR optimization (e-mail blasts, mass training sessions, etc). In our experience, redesigning clinical workflow in tandem with efforts to reduce the clinician EHR burden was critically important to improving clinician satisfaction with the EHR. Workflow changes sometimes necessitated retraining of clinicians and clinic staff on EHR tools. However, workflow changes almost always required effective interpersonal communication, task delegation, and the creation of clinic standard policies. With the Sprint method, once a new workflow was agreed upon, the clinic could implement the change immediately. Coordinated and rapid changes in clinical workflow and improved EHR efficiency were a source of substantial clinician satisfaction. For example, some clinics implemented a standard prescription renewal process for 90 days and 3 additional refills, reducing unnecessary variation. One clinic developed a standard previsit online patient questionnaire about neurological disorders to be used for every visit.
      Several other elements of the Sprints intervention were critical to its success. Chief among these was having a physician informaticist leader on the Sprint team to efficiently translate clinician requests for EHR changes into technical and/or training solutions, a nurse informaticist to focus on staff and team workflows, a dedicated project manager, and consistent Sprint team members whose collective knowledge and effectiveness increased with each Sprint. Having the entire Sprint team on-site in the clinic was also invaluable, because the team could huddle and make decisions quickly. Equally important was the face-to-face involvement of clinic leaders and clinician superusers in the process. Sprint was not a spectator sport; it was an all-hands-on-deck effort, and the more the clinic leaders and clinicians participated, the more each clinic benefited. The short time frame and rapid tempo of Sprint showed clinicians that improvements were being made each day, which helped to build confidence in the process. Some clinician superusers continued to be EHR experts for their colleagues long after the Sprint process concluded, an ideal situation for long-term improvement. Clinicians often believe that the main problem with practice efficiency is a poorly designed EHR with insufficient specialty tools. However, by the end of Sprint, we estimated that 80% of the improvement in clinician experience resulted from training clinicians on existing tools and redesigning teamwork, with only about 20% coming from newly customized EHR tools.
      Executing Sprint was not without challenges. There is always organizational pressure to increase clinical productivity and see more patients, which can be at odds with efforts to reduce clinician burnout. We were able to convince leaders to protect clinician time for participation in Sprint, because busy clinicians without protected time are unlikely to participate. It is important that leaders do not expect an increase in patient volume as a direct result of Sprint. In addition, despite these measures, about one-third of clinicians did not participate adequately in Sprint. It is possible that these clinicians opted out because they judged themselves to already be using the EHR optimally. However, on the basis of our in-clinic observations, we believe that nonparticipators may be so burned out and disengaged that offering ways to improve practice efficiency is insufficient. Complementary approaches, such as building a “culture of wellness” or improving “personal resilience,” may also be needed.
      • Bohman B.
      • Dyrbye L.
      • Sinsky
      • et al.
      NEJM Catalyst. Physician well-being: the reciprocity of practice efficiency, culture of wellness, and personal resilience.

      Cost and Benefit

      The team of 11 costs $1.2 million annually, including one internal medicine physician salary at about $280,000 including benefits. With a full schedule, a Sprint team can accomplish seventeen 2-week Sprints annually. The team can engage 30 clinicians per Sprint, or about 500 clinicians per year, which results in approximately $2400 spent per clinician. Also, for every clinician in Sprint, we request 7 hours protected time. Using 52 weeks per year and 40 hours a week, a full-time clinician spends 2080 hours per year. Per clinician at $280,000, 7 hours loss of productivity is $942, or about $1000. Aggregating, Sprint may cost an organization $3400 per clinician. Using these assumptions, cost per 500 clinicians in a 1-year Sprint may be $1.7 million.
      Using assumptions from the AMA STEPS forward physician burnout calculator,
      • Shanafelt T.
      • Goh J.
      • Sinsky C.
      The business case for investing in physician well-being.
      AMA STEPSforward
      Organizational changes lead to physician satisfaction.
      if 12.5 physicians out of 500 burn out per year, an intervention may reduce burnout by 20% or by 2.5 physicians, and cost of replacing a physician is $1 million, it works out to be $2.5 million cost avoidance. Measured another way, $3400 per clinician per year could be offset by each clinician seeing 1 more patient every 2 weeks.

      Limitations

      This intervention does have several limitations. Lacking a standardized instrument to assess the EHR user experience, we chose to use NPS methodology and a short survey to assess clinician satisfaction with Sprint and with the EHR. However, neither the NPS instrument nor the survey questions have been validated for use in this setting. A second limitation is that the Sprints intervention constantly evolved over time as the Sprint team strove to best meet the needs of each individual clinic. Thus, not all clinicians received exactly the same Sprints intervention or the same version of the survey. Furthermore, we did not identify and survey a comparable control group. Lastly, Sprint was carried out at a single US health care organization. In the 6 participating clinics, academic clinicians outnumbered community clinicians, and the findings may not be generalizable to other health care settings.

      Conclusion

      The Sprint approach to EHR optimization represents a viable option for improving teamwork, reducing clinician EHR burden, and improving clinician satisfaction with the EHR. In this brief intervention, emotional exhaustion from burnout did not change. The Sprint optimization effort was so well received by clinicians that funding for a second full Sprint team has been allocated to increase the pace of this work. Additional organizational experience, methodological refinements, and the use of validated assessment methods are needed to determine whether the Sprints intervention can be applied to solve other challenges associated with EHR-enabled patient care.

      Acknowledgments

      We acknowledge the invaluable assistance of Esther Langmack, MD, for her editorial assistance and of Hillary Lum, MD, Christine Sinsky, MD, and John Steiner, MD, for their close reading of the manuscript.

      Supplemental Online Material

      Supplemental material can be found online at: http://www.mayoclinicproceedings.org. Supplemental material attached to journal articles has not been edited, and the authors take responsibility for the accuracy of all data.

      References

        • DesRoches C.M.
        • Charles D.
        • Furukawa M.F.
        • et al.
        Adoption of electronic health records grows rapidly, but fewer than half of US hospitals had at least a basic system in 2012.
        Health Aff (Millwood). 2013; 32: 1478-1485
        • Adler-Milstein J.
        • DesRoches C.M.
        • Furukawa M.F.
        • et al.
        More than half of US hospitals have at least a basic EHR, but stage 2 criteria remain challenging for most.
        Health Aff (Millwood). 2014; 33: 1664-1671
        • Arndt B.G.
        • Beasley J.W.
        • Watkinson M.D.
        • et al.
        Tethered to the EHR: primary care physician workload assessment using EHR event log data and time-motion observations.
        Ann Fam Med. 2017; 15: 419-426
        • Shanafelt T.D.
        • Hasan O.
        • Dyrbye L.N.
        • et al.
        Changes in burnout and satisfaction with work-life balance in physicians and the general US working population between 2011 and 2014.
        Mayo Clin Proc. 2015; 90 ([published correction appears in Mayo Clin Proc. 2016;91(2):276]): 1600-1613
        • Shanafelt T.D.
        • Dyrbye L.N.
        • Sinsky C.A.
        • et al.
        Relationship between clerical burden and characteristics of the electronic environment with physician burnout and professional satisfaction.
        Mayo Clin Proc. 2016; 91: 836-848
        • Friedberg M.W.
        • Chen P.G.
        • Van Busum K.R.
        • et al.
        Factors affecting physician professional satisfaction and their implications for patient care, health systems, and health policy.
        Rand Health Q. 2014; 3: 1
        • Bodenheimer T.
        • Sinsky C.
        From triple to quadruple aim: care of the patient requires care of the provider.
        Ann Fam Med. 2014; 12: 573-576
        • Dewa C.S.
        • Jacobs P.
        • Thanh N.X.
        • Loong D.
        An estimate of the cost of burnout on early retirement and reduction in clinical hours of practicing physicians in Canada.
        BMC Health Serv Res. 2014; 14: 254
        • Noseworthy J.
        • Madara J.
        • Cosgrove D.
        • et al.
        Health Affairs Blog. Physician burnout is a public health crisis: a message to our fellow health care CEOs.
        • Shanafelt T.
        • Goh J.
        • Sinsky C.
        The business case for investing in physician well-being.
        JAMA Intern Med. 2017; 177: 1826-1832
        • Bohman B.
        • Dyrbye L.
        • Sinsky
        • et al.
        NEJM Catalyst. Physician well-being: the reciprocity of practice efficiency, culture of wellness, and personal resilience.
        • Scrum.org
        The Home of Scrum: An Advanced Scrum Master Class.
        https://www.scrum.org
        Date accessed: July 16, 2017
        • Reichheld F.F.
        The one number you need to grow.
        Harv Bus Rev. 2003; 81: 46-54, 124
        • Qualtrics
        What is Net Promoter Score (NPS)?.
      1. Insights From Analytics. Top 10 U.S. Net Promoter Scores (NPS) for 2013.
        • Net Promoter System®
        Companies that use net promoter.
        • West C.P.
        • Dyrbye L.N.
        • Sloan J.A.
        • Shanafelt T.D.
        Single item measures of emotional exhaustion and depersonalization are useful for assessing burnout in medical professionals.
        J Gen Intern Med. 2009; 24: 1318-1321
        • Lyon C.
        • English A.F.
        • Chabot Smith P.
        A team-based model that improves job satisfaction.
        Fam Pract Manag. 2018; 25: 6-11
        • AMA STEPSforward
        Organizational changes lead to physician satisfaction.

      Linked Article

      • In the Limelight: May 2019
        Mayo Clinic ProceedingsVol. 94Issue 5
        • Preview
          This new monthly feature highlights four articles in the current print and online issue of Mayo Clinic Proceedings. These articles are also featured on the Mayo Clinic Proceedings' YouTube Channel ( https://mayocl.in/2U1xy5H ), and may be also discussed by an accompanying editorial.
        • Full-Text
        • PDF