Transcription

Adapting the EHR Scribe Model toCommunity Health Centers:The Experience of Shasta Community Health Center’s PilotPREPARED FOR:PREPARED BY:Kim Ammann HowardKris HeléNadia SalibiSheila WilcoxMoriah Cohen

Adapting the EHR Scribe Model toCommunity Health Centers:The Experience of Shasta Community Health Center’s PilotPREPARED FOR:PREPARED BY:Kim Ammann HowardKris HeléNadia SalibiSheila WilcoxMoriah Cohen

Table of ContentsAdapting the EHR Scribe Model to Community Health Centers . 1Introduction . 1Evaluation Findings . 1Looking Forward . 9Appendix A: Methods . A1

PrefaceACKNOWLEDGEMENTSBTW informing change (BTW) is grateful to the staff at Shasta Community Health Center (SCHC) for theircommitment and contribution to this evaluation. In particular we would like to thank SCHC staff members whoprovided their guidance on the evaluation throughout the study period. We offer a special thanks to CharlesKitzman, Chief Information Officer, Julie Johnston, Director of Finance, and Alexis Parsons, Director of HealthInformation Services, for their extraordinary assistance with the various aspects of data collection. We are alsograteful to the clinicians, scribes, patients and other staff who participated in the evaluation and contributed theirtime and insights. We also want to acknowledge Blue Shield of California Foundation for their support of thiswork.For more information about this project, please contact Charles Kitzman at 530.246.5955 [email protected] more information about this evaluation, please contact Kim Ammann Howard, Director of Evaluation &Organizational Learning at 510.665.6100 or [email protected] BTW informing changeAt BTW we are driven by our purpose of informing change in the nonprofit and philanthropic sectors. We partnerwith our clients to improve their effectiveness and build a culture of learning and continuous improvement. Weproduce high-quality, easy-to-understand products that present useful information designed to be readily appliedto practice. Our information-based services focus on the fields of health, education, youth engagement andphilanthropy.To find out more about BTW and our services, visit www.btw.informingchange.com.i

INTRODUCTIONShasta Community Health Center (SCHC) was established in 1988 in response to the lack of health care servicesavailable for the underserved community in Redding, California and its surrounding areas. With a mission toprovide comprehensive, high quality, efficient and effective health care services, SCHC continues to enhance itscapacity to meet the needs of a growing patient population. Four years ago, in 2008, this included the adoption ofElectronic Health Records (EHR), putting SCHC at the forefront of integrating systems to promote efficiency forits clinical staff and to provide high quality care for its patients. An unanticipated outcome of EHR adoption hasbeen the extra burden of work placed on SCHC clinicians, which to different extents negatively affects theirproductivity and satisfaction. To address these challenges, in January 2011, SCHC piloted an EHR scribe model.In brief, this involves a scribe being present during the patient visit to enter the appropriate information into theEHR, as dictated by the clinician.While this practice is fairly prevalent in emergency rooms and to a lesser degree, in specialty care settings, itremains uncommon in community health centers. This provided SCHC with a particularly timely opportunity toassess how scribes could mitigate the above stated challenges as well as applicability of the model to communityhealth centers more broadly. SCHC, with assistance from Blue Shield of California Foundation, contracted withBTW informing change (BTW) to assess the model’s effectiveness. The evaluation focused on answering thefollowing questions: Does clinician satisfaction improve due to having a scribe?Does patient satisfaction improve due to having a scribe?Does patient chart documentation improve due to having a scribe?What impact does the EHR scribe model have on patient visit efficiency and flow?What is the fiscal impact of the EHR scribe model?What clinician and scribe characteristics are important for the success of the EHR scribe model?To answer these questions, primary data were collected during the study period, July through October 2011, fromclinicians who worked with a scribe (n 6), scribes (n 8), patients who experienced a visit with a scribe (n 221),and key SCHC staff involved in the EHR scribe implementation. Secondary data included a review of patient chartdocumentation and EHR data extraction for the study period as well as the baseline comparison period, Julythrough October 2010. For the EHR-related data, a comparison group of clinicians who worked with a scribe 14days or less (n 6) were chosen based on similar specialty, hours worked and experience. A more in depthdescription of the methods and key limitations can be found in Appendix A.EVALUATION FINDINGSIn this section, we provide a summary of key findings for each evaluation question above. Where multiple datasources exist for a specific evaluation topic area, findings were analyzed jointly to provide a more completepicture.Does clinician satisfaction improve due to having a scribe?Overall clinicians are more likely to report being “much more satisfied” or “more satisfied” when working with ascribe as compared to when not working with a scribe (Exhibit 1). The areas with the greatest improvements insatisfaction are the timeliness of chart notes, overall time spent on chart notes and overall job satisfaction. While afair percentage of clinicians report levels of satisfaction remaining the same regardless of having a scribe, only oneclinician reports feeling “less satisfied” in one area: the overall amount of time spent on chart notes.1

Exhibit 1Clinicians’ Satisfaction When Working with a Scribe Compared to Not Workingwith a Scribe(n 5)Clinicians’ Satisfaction Level*Same level ofsatisfactionMore satisfiedMuch moresatisfiedAbility to stay on schedule overall Timeliness of chart notes Less satisfiedTime spent on chart notes overall Overall accuracy of chart notes Overall workload Overall job satisfaction *Each ―x‖ represents one clinician.Clinicians were also asked about the appropriateness of scribes’ skills andqualities. With one exception, all clinicians note that they feel “very satisfied” or“somewhat satisfied” with the following scribe attributes: Eagerness to helpHard workingAbility to get along with others, particularly the clinician’s clinic teamInteraction with patientsService focusClinicians report the same levels of satisfaction as noted above with how well theyand scribes work together and how well scribes meet their needs.“Decreasing stressin the clinical areais very important.The scribe doesthis by having mynotes almostcompleted.”—Shasta clinicianConcerns About Working with a ScribeClinicians were asked to reflect on their concerns about working with a scribe. Exhibit 2 highlights initial concerns(i.e., clinicians’ concerns prior to working with a scribe) reported by at least three or more clinicians. These rangedfrom the time it would take to train and get used to working with a scribe to insufficient skills among scribes to thenegative impacts on their patients of having a scribe in the exam room. After working with a scribe, however, mostclinicians no longer report having these concerns. For the further development of scribes, clinicians recommendadditional training in medical terminology and documentation, appropriate medical coding and note taking.Other concerns that were voiced initially by one or two clinicians (e.g., scribe-clinician relationship, reliance onscribes, not adequately learning how to use EHR) were no longer a concern after working with a scribe.Clinicians vary in the number of patient visits they think it takes to become comfortable working with a scribe.The mean was 50 patient visits with a range of 20 to 100 patient visits.2

Exhibit 2Changes in Clinicians’ Concerns About Working with a Scribe*(n 5)Prior to Working witha ScribeAfter Workingwith a ScribeTime or effort to train scribes Time or effort to get used to working withscribes Patients might be uncomfortable with ascribe in the room Scribes might lack sufficient technical orprofessional skills Scribes could negatively affect clinicians’relationship with patients *Each ―x‖ represents one clinician.Impact of a ScribeIn reflecting on their overall work experience, four of the fiveclinicians feel confident that working with a scribe will increase thelength of time they stay in clinical care. This included the most seniorclinicians.1 They point to feeling less stressed and overwhelmed and adecrease in their workload. In fact, all clinicians note that the length oftheir work day has either decreased or“Having a scribe remained the same since working with ascribe. As one clinician reflects, “I’mis the differenceworking 10 hour days instead of 12 hourbetween feelingdays.”hopeless andoverwhelmedand feeling likeit’s a doable joband verysatisfying.”—Shasta clinician1Other benefits clinicians note include thepositive impact of having another personin the room on patient management andmore “face time” with patients. Notes oneclinician: “I am able to listen to patientsand not worry that I am writing everythingdown . Also having another set of ears ishelpful for difficult or uncooperativepatients.”SUMMING IT UP:SATISFACTION WITHSCRIBESClinicians’ Satisfaction Clinicians report greater satisfactionwith the time spent on chart notes,the accuracy of chart notes andtheir overall workload. After working with a scribe,clinicians’ initial concerns tend todiminish. Almost all clinicians think that theywill stay in clinical care longer as aresult of having a scribe.Patient Satisfaction Most patients did not voice anyconcerns about having a scribe inthe room during their visit.The average age of clinicians was 52 with a range of 36 to 69 years.3

Does patient satisfaction improve due to having a scribe?“I like it better with aWhen provided with the opportunity to voice concerns about having a scribein the room during their clinic visit, almost all patients (90%) say that theydid not have concerns. One patient notes, “I agree with having a scribe in theroom if it helps the doctor.” Those with concerns report issues related to lackof privacy, confidentiality and discomfort with another person’s presenceduring the exam. This sentiment was expressed by one patient who stated, “Itis kind of awkward to have a third party listening.” Patients could always askfor a scribe to leave a room. Patient satisfaction did not vary by age, ethnicity,gender or type of patient visit (i.e., regular checkup, acute or chronic healthissue).scribe in the roombecause the doctor isable to spend moretime with me ratherthan entering theinformation into thecomputer.”—Shasta patientPatient & Clinician Satisfaction with ScribesOverall, most clinicians and patients report high levels of satisfaction with the EHR scribe model. They were askedabout different aspects of communication between the clinician and patient as well as their overall experience witha scribe in the room. Across these measures, a slightly greater percentage of clinicians note higher levels ofsatisfaction as compared to patients. The one exception is “how well the patient can communicate with theclinician” for which a slightly greater percentage of patients report higher levels of satisfaction (Exhibit 3).Exhibit 3Clinicians’ & Patients’ Satisfaction with a Scribe in the Room*P Patient (n 217–220); C Clinician (n 5)CommunicationHow clearly clinician canexplain things to patientHow well clinician cancommunicate with patient1%59%40%1%20%3%How well clinician can listenHow well patient cancommunicate with clinicianOverall Experience20%5%20%59%59%40%60%80%80%64%60% Less satisfied38%38%PCPCPC31% P20%CSatisfaction with overallexperience during patient visitsThe quality of care clinician canprovide during patient visitsComfort feltduring visitHow well clinican canmeet patients' needsHow much time clinician canspend with patients Same level of 0%80%36%PC34%PC31%PC80%60%40%PC37%60%PC75% More satisfied*Some rows do not add up to 100% due to rounding.4

Does patient chart documentation improve due to having a scribe?The quality and accuracy of patient chart documentation improved among the study group clinicians during thestudy period as compared to the baseline period for the three key measures described below: Chief Complaint (CC): Description of the symptom, problem, condition, diagnosis, physicianrecommended return or other presenting reason for a medical encounter as usually described by thepatient.Evaluation and Management Coding (E/M): A medical billing process that doctors must use to bereimbursed by Medicare, Medicaid programs or private insurance for a patient encounter.International Classification of Diseases and Related Health Problems (ICD-9): A medicalclassification that provides codes to classify diseases and a wide variety of signs, symptoms, abnormalfindings, complaints, social circumstances and external causes of injury or disease.The accuracy of CC and ICD-9 coding increased 10 percentage points from 88% to 98% for CC scores and 87% to97% for ICD-9 scores. The greatest improvement took place with E/M coding scores with a 17 percentage pointincrease in accuracy from 61% to 78% (Exhibit 4).Exhibit 4Improvements in Quality & Accuracy of Chart DocumentationAmong Clinicians Before & After Using a Scribe(n 6)Percentage Change in Quality &Accuracy of Chart Documentation98%90%88%97%87%78%70%61%50%Baseline period (July–Oct 2010)E/M Coding ScoreICD-9 ScoreStudy period (July–Oct 2011)CC ScoreWhat impact does the EHR scribe model have on patient visit efficiency and flow?To assess the impact of having a scribe on patient visit efficiency and flow, the number of additional patientencounters was calculated2 for clinicians in the study group as well as those in the comparison group. The studygroup saw, on average, an additional 54 patients during the study period as compared to the baseline period. Thecomparison group saw, on average, an additional 29 patients during the study period as compared to the baselineperiod (Exhibit 5). Based on these calculations, the EHR scribe model appears to have positively impacted patient2Additional patient encounters is based on the difference between the 2011 actual patients and 2011 expected patients; the latter of which isdetermined by 2011 clinic hours and 2010 productivity.5

flow at SCHC with clinicians who use a scribe on average seeing 25 more patients over the four month period, ascompared to the comparison group.Exhibit 5Average Additional Patient Encounters Per Clinician in theStudy Group & the Comparison GroupAdditional PatientEncounters60545040293020100Calculated based on clinic hoursStudy Group (n 6)Comparison Group (n 6)What is the fiscal impact of the EHR scribe model?Clinician productivity and clinician revenues were used to assess the fiscal impact of the EHR scribe model. Theaverage productivity rate3 among clinicians in the study group was lower during the baseline period as well as thestudy period as compared to the comparison group. However, the average productivity rate for clinicians in thestudy group increased at twice the rate (6%) as that for the comparison group (3%) (Exhibit 6).Exhibit 6Changes in the Average Productivity Rate Among Clinicians in theStudy Group & the Comparison GroupChange in Average 31.21.11Baseline period (July–Oct 2010)Study Group (n 6)31.911.85Study period (July–Oct 2011)Comparison Group (n 6)Average productivity rate is based on number of clinic encounters divided by clinic hours.6

To evaluate the impact of the EHR scribe model on clinic revenues, average revenue per clinician4 and the relatedcosts for a scribe5 were computed to calculate the average net revenue per clinician for both the study group andcomparison group. The study group demonstrated higher average revenues per clinician as compared to thecomparison group, 7,551 and 4,047 respectively, due to the greater number of additional patient encounters(Exhibit 7). However, the average cost of a scribe in the study group, 5,035, offset this gain and resulted in loweraverage net revenues per clinician than in the comparison group ( 2,517 versus 3,802 respectively). Ultimately,while the use of scribes resulted in a gain in revenue for the study group, because the cost of a scribe demandsadditional funds, the study group did not attain as much of a gain in net revenue as the comparison group.Exhibit 7Average Net Revenue Per Clinicianfor the Study Group & Comparison GroupStudy Group (n 6)AverageRevenueIncreasePer Clinician 7,551Cost of aScribe PerClinician 5,035AverageNetRevenueIncreasePer Clinician 2,517Comparison Group (n 6)AverageRevenueIncreasePer Clinician 4,047Cost of aScribe PerClinician 245AverageNetRevenueIncreasePer Clinician 3,802What clinician and scribe characteristics are important for the success of the EHR scribemodel?Clinicians, scribes and key SCHC staff involved in the model’s implementation provided feedback on whatcontributes to a successful EHR scribe model. They were asked specifically about the qualities of clinicians thatbest compliment scribes, characteristics of a successful scribe and important considerations for a successfulscribe-clinician match.45Average revenue is calculated by multiplying additional patient encounters by 140.This includes an hourly wage of 13 for scribes multiplied by 1.3 to account for administrative costs multiplied by scribe hours billed. Thisdoes not include the initial cost of scribe training or the ongoing cost of scribe development.7

Key Characteristics of Clinicians Best Suited to Work with an EHR ScribeBased on their experiences working with clinicians, scribes note a few important characteristics that make aclinician best suited to work with a scribe, including adaptability, good communication skills and anunderstanding of scribes’ role and how to best utilize them during a patient visit. Staff overseeing projectimplementation describe the importance of clinician readiness for a scribe, which includes a willingness to investupfront time to get the scribe up to speed, learn a new way of working, and delegate. Some clinicians and staffnote the value of the EHR scribe model for specific types of patient visits such as preventive, urgent care andspecialty care. They raise the point that patients with more complicated health issues may be too difficult for ascribe without medical training.Some staff reflect on the importance of having at least one clinician advocate to encourage those who may beresistant to the model. While a number of SCHC clinicians were resistant to working with a scribe when theproject began, over time, an increasing number of clinicians have been requesting a scribe. This, in large part, hasbeen a result of clinician peers talking about their positive scribe experience and the tangible aforementionedbenefits. It also has been invaluable to have a highly competent lead scribe with strong interpersonal skills andsensitivity to clinicians’ needs who can work with them throughout the process.Key Characteristics of a Successful ScribeClinicians identify what they think are the most important technical skills and capabilities for a scribe to besuccessful. Most commonly they note knowledge of medical terminology, spelling skills and experience with EHR.Less commonly noted are backgrounds or skills related to college education, experience in the health care fieldand communication skills. Scribes echo similar sentiments, adding the need for scribes to be able to think quicklyand deal with stressful patients. Other staff also reflect on the importance of scribes who are friendly andfascinated with people and learning.While staff reflect on the benefits of being able to hire scribes with college degrees, they also reflect on the value ofscribes who may not have such degrees but who come to the position with prior exposure to clinical care andknowledge of medical terminology. Clinicians, scribes and other staff highlight the benefit of hiring scribes on themedical pathway (e.g., someone who plans to pursue clinical training as a nurse, physician’s assistant or doctor)and the rich experience that scribing provides. Some remark, however, that while this may be helpful in the shortrun, it could mean shorter tenure at the health center.Establishing a Successful Scribe-Clinician MatchClinicians and scribes vocalize similar thoughts on what makes a successful scribe-clinician match. They note theimportance of preparing both scribes and clinicians to work together. For scribes, this involves a formal andintensive training that includes the following: 1) a general introductory training for all scribes, 2) an opportunityto shadow the lead scribe and matching clinician to watch “scribing in action,” and 3) an opportunity to practice“scribing” under the watchful eye of the same scribe-clinician pair and receive constructive feedback. Next,customized trainings take place to facilitate a good fit between an individual scribe and clinician. The lead scribeworks with the clinician for a short time to inform decisions about the most appropriate scribe-clinician match.Once the scribe-clinician match is made, it is important to allow adequate time for the scribe to work with theclinician to learn specific preferences, which vary among SCHC clinicians. As one scribe describes, “Every clinicianis so different. We really have to adapt to their style . Once you get that down, you are good.” The importance ofdeveloping a consistent and stable relationship is a strong theme addressed by both clinicians and scribes; such arelationship facilitates trust, strong communication and good rapport. Sharing information among scribes aboutclinician preferences and experiences also has been helpful when a scribe-clinician match shifts temporarily orpermanently (e.g., due to a scribe being sick or leaving their position).8

LOOKING FORWARDThis pilot of the EHR scribe model sets forth the foundation for a more informed discussion on the continuousefforts to advance and develop best practices in a health care setting. With EHRs becoming the standard ofpractice in the clinical environment and the increase in popularity of scribes, the marriage of the two seems like anatural next step. This study sheds light on a relatively new approach for integrating the EHR scribe model intocommunity health centers. In spite of the evaluation’s limitations due to the EHR scribe implementation being inits early stages and the small number of participating clinicians, the model builds on positive anecdotal stories andshows promise in a number of key impact areas. Overall, there was a neutral or positive impact of having a scribefor both clinicians and patients. Improvements in patient chart documentation, which is critical for patient care,safety and reimbursement, occurred across all measures, with small percentage changes having practicalimportance. While the impact of additional patient encounters and net fiscal gains may not be as high in the studygroup as desired, these impacts may improve over time, among individual SCHC physicians currently using ascribe and as other clinicians begin to do so. It also will be important to continue to assess which clinicians arebest suited for a scribe, considering such factors as baseline productivity, newness to the health center andtimeframe in which the highest productivity levels are realized. The extent to which the role of scribes can bebroadened should also be considered. For example, scribes could assume duties typically undertaken by medicalassistants, which do not require medical licensure (e.g., contacting patients with normal lab results, schedulingreferrals, preparing patient tracking reports). The collection of additional data over time will allow for moredefinitive conclusions including the practice of providing a scribe to every clinician in community health centersas compared to matching scribes to those clinicians who can become even more efficient. It will also be helpful toassess the effectiveness of all of these study measures at other community health centers to get a greater sampleboth within and outside of SCHC.9

APPENDIX AEvaluation MethodsDATA COLLECTION METHODSBTW worked with Shasta Community Health Center (SCHC) staff to develop a robust evaluation plan thatintegrates a mixed methods approach of data collection strategies and incorporates feedback from all populationsimpacted by the EHR scribe model. The table below summarizes the evaluation methods utilized, includingtimeframe, type of data collection and sample for each key outcome measure. Subsequently, we provide someadditional detail about each method.OutcomeMeasureTimeframeType of Data CollectionSample Type and SizeSatisfaction:Assess satisfactionwith the EHRscribe model forclinicians andpatients.Jul–Nov 20111 clinician focus group3 participants*Clinician satisfaction survey5 participantsPatient satisfaction survey221 participantsChartDocumentation:Review patientcharts to assessaccuracy andquality.Jul–Oct 2011**Chart reviews6 participantsEfficiency:Assess EHR datafor billable clinichours.Jul–Oct 2011**EHR data extraction12 participants***FiscalContribution:Assess EHR datafor averagerevenues perclinician.Jul–Oct 2011**EHR data extraction12 participants***LessonsLearned:May–Nov 20112 scribe focus groupsJuly: 6 participantsProvide insight onimportantcharacteristics ofclinicians who usescribes, scribesand the scribeclinician match.November: 6 participantsPeriodic reflections with key staffinvolved in modelimplementation****Average of 5 per meeting* One of these participants was not included in other clinician data collection sources.** Baseline data also was extracted for these measures from the July–October 2010 time period.*** This consists of a study group of six SCHC clinicians and a comparison group of six SCHC clinicians.**** Data from the clinician focus group and satisfaction survey also was utilized.A1

APPENDIX AClinicians included in the study group had worked with a scribe for at least 50 patient visits during the six-weekperiod prior to the start of the study period, July–October 2011, and continued to work with scribes during thisperiod. Three clinicians participated in an in-person focus group at the onset of their scribe experience in July2011. Five of the six clinicians in the study group completed a short online clinician satisfaction survey in October2011 (the sixth clinician no longer worked at the clinic at the time of survey administration). SCHC clinicians wereeligible to be in the comparison group if they had a history of working with a scribe for 14 days or less sinceNovember 2010. Each clinician in the study group was matched with a clinician based on specialty, hours workedand clinical experience. In general clinicians in the study group had lower baseline productivity as compared tothose in the comparison group.Patients were administered a patient survey by telephone by an SCHC staff person during July–November 2011.Patients were eligible to take the survey if they were 18 years of age or older, their primary language was Englishand they had been examined by an SCHC clinician in the study group1 within the last three weeks with a scribe inthe room as well as the same clinician within a 14 month period without a scribe being present in the room. On aweekly basis, SCHC staff generated a list of eligible patients to contact and attempted to reach approximately 5patients per clinician each week, with an ultimate goal of 50 patients per clinician. Patients could be contactedmultiple times within their three week eligibility period; however, they could be surveyed only once per clinician.For two clinicians the goal of 50 was not reached due to eligibility criteria for patient inclusion; forty-five surveyswere completed for one of these clinicians and 18 for the other clinician.Survey findings show that the responding patients closely resemble the demographics at SCHC. A total of 221patients were surveyed on their experience with their clinician when having a scribe in the room. The compositionof the patient population was predominately white (82.5%) and had a small Hispanic/Latino make-up (5.6%);majority female (65.5%); and ranged in age from 18–88 years (mean of 53.41 years). Patients reported seeing theirclinician an average of 5.09 times per year with the number of visits ranging from 2 to 54. While 68.3% reportedthat their clinic visit was for some form of a chronic or acute health issue, 31.7% identified the main reason fortheir visit as a regular checkup.Scribes participated in two focus groups to discuss their experiences as a scribe. The first took place in person inJuly 2011 and the second by phone in November 2011. Four of the six scribes in each focus group participated inboth. Three of the six scribes in the first focus group and five of the six in the second focus group are the scrib

Secondary data included a review of patient chart documentation and EHR data extraction for the study period as well as the baseline comparison period, July through October 2010. For the EHR-related data, a comparison group of clinicians who worked with a scribe 14 days or less (n 6) wer