Adm Policy Ment Health DOI 10.1007/s10488-015-0648-7
ORIGINAL PAPER
Progress Monitoring in an Integrated Health Care System: Tracking Behavioral Health Vital Signs Bradley Steinfeld1 • Allie Franklin2 • Brian Mercer1 • Rebecca Fraynt1 Greg Simon3
•
Springer Science+Business Media New York 2015
Abstract Progress monitoring implementation in an integrated health care system is a complex process that must address factors such as measurement, technology, delivery system care processes, patient needs and provider requirements. This article will describe how one organization faced these challenges by identifying the key decision points (choice of measure, process for completing rating scale, interface with electronic medical record and clinician engagement) critical to implementation. Qualitative and quantitative data will be presented describing customer and stakeholder satisfaction with the mental health progress monitoring tool (MHPMT) as well as organizational performance with key measurement targets. Keywords Progress monitoring Behavioral health Integrated health care system
Introduction It is well documented that measurement of performance and providing feedback, often referred to as progress monitoring (Bickman et al. 2011; Lambert 2010), is foundational to improvement in mental health care. Institute of Medicine of the National Academies (2001, 2006) recommended that both mental health and health care providers regularly use valid and reliable patient & Bradley Steinfeld [email protected] 1
Group Health Cooperative, 950 Pacific Ave Suite 900, Tacoma, WA 98402, USA
2
Optum Health Care, Tacoma, USA
3
Group Health Research Institute, Seattle, USA
assessment instruments to measure treatment progress and outcomes. Despite the research evidence that progress monitoring improves treatment outcomes (Lambert 2010), as well as the development of a number of mental health progress monitoring systems (Bickman et al. 2011; Chorpita and Daleiden 2010; Duncan et al. 2010; Lambert 2010), the majority of clinicians do not use these tools as a routine part of their practice. Many of the barriers to implementation are similar to those identified in implementation of evidence based practices (EBP). These include lack of training, staff engagement, organizational readiness and implementation structures (Fixsen et al. 2005; McHugh and Barlow 2012; Steinfeld et al. 2014). In addition, there are a number of barriers as it pertains to clinician attitudes including time and cost to use tool (Jensen-Doss and Hawley 2010), clinician hesitance to adopt computer-based technology (Lambert 2010), clinician belief they are in a better position to judge patient progress than the patients themselves and process barriers such as administrative complexity and cost (Lambert 2010). As behavioral health care is increasingly integrated within health care delivery systems (McDaniel and Degruy 2014), there are additional implementation challenges. Current behavioral health progress monitoring systems are typically only familiar to mental health practitioners and are tracked using stand-alone platforms that do not integrate easily with fully integrated electronic medical records or chronic disease management programs (Steinfeld and Keyes 2011). This article will describe the experience of implementing a Mental Health Progress Monitoring Tool (MHPMT) in an integrated health care delivery system. An overview of the setting in which this work occurred will be provided along with the contextual factors which led this
123
Adm Policy Ment Health
organization to develop a MHPMT system. The process of implementation will be described including the measurement selection, administrative, and training processes, as well as how pilot implementation influenced full-scale implementation. Full-scale implementation will be described with a focus on staff engagement, integration of MHPMT with electronic medical record (EMR), organizational sponsorship, and measurement targets. Results will be shared which will include patient and staff qualitative feedback, rates of MHPMT usage and patient outcomes. Finally, critical factors leading to successful implementation, as well as continued challenges, will be shared.
The Setting Group Health Cooperative is a consumer-governed, nonprofit, health care system that provides coverage and heath care for approximately 650,000 people in Washington and Idaho. Medical care is provided in 26 medical center and clinics by over 1000 medical providers. The Behavioral Health Service Department at Group Health Cooperative provides coverage, mental health and chemical dependency care for 17,000 people per year via its seven model mental health clinics, as well as its contracted network. The target audience implementing the progress monitoring tool consisted of the mental health providers in these seven staff model clinics (approximately 125 staff), composed of master level clinicians, psychologists, psychiatrists and psychiatric nurses.
Getting Started/A Case for Change For a number of years, Group Health had been focused on improving patient satisfaction as measured by patient response to surveys assessing satisfaction with overall treatment (Crosier et al. 2012). Some initial work indicated that regular use of progress monitoring tools such as the Burns Mood Enhancement Scale (Burns 1997) and Outcome Rating Scale (Miller et al. 2003), increased patient satisfaction scores. However, this initiative was not continued as there were a number of barriers with these progress monitoring tools, which included clinician dissatisfaction with items on the scales, difficulty tracking data over time, and difficulty consistently administering the scales to patients.
Pilot Program Subsequently, one of the mental health clinics had low patient satisfaction survey scores, and was interested in again piloting whether systematic use of a progress monitoring tool
123
could improve patient satisfaction scores. This pilot program involved one group health behavioral health clinic made up of 13 clinicians (2 psychiatrists, 2 psychiatric nurses, 2 psychologists, and 7 master level therapists) all of whom participated in the progress monitoring pilot. In order to address the difficulties encountered during the initial, smallscale study, a third party vendor was solicited to implement a MHPMT. A key issue was deciding what items should be on the progress monitoring measure. As Scott and Lewis (2015) have noted, an important consideration is determining whether the items should be idiographic (i.e., tailored to the patient’s specific problem) or nomothetic (i.e., standardized self-report measures that can be administered consistently over time). Based upon our initial experience and in consultation with the third party vendor, we decided to use a nomothetic approach which would simplify administration (same measure for each patient at each visit) and we used items that the vendor had identified were sensitive to change in a wide variety of clinical conditions (i.e., global distress) and were also psychometrically valid. Severity adjusted effect size, a standardized method for determining change that takes into account case mix and symptoms severity (Brown and Minami 2010) was the measure used by the vendor to determine clinical significance of change across a patient population. The vendor also had a free standing web site which included raw scores on an individual basis, a graphical representation of the change in raw scores for individual patients, and severity effect change scores for population of patients on the provider, clinic, and organizational level. To assure that patients consistently completed the MHPMT, support staff distributed the forms to all adult patients over the age of 12 at all visits. Patients gave the completed forms to their providers, who then returned the forms to the front desk staff. These forms were then faxed to the third-party vendor on a daily basis. Clinicians were also trained to use the MHPMT through didactic presentations. Information was provided on theoretical/clinical rationale for progress monitoring scales, how to interpret the scales and discuss the results with patients, and how to access the website of the third party vendor. Follow-up consultations occurred on a monthly basis for 2–3 months post-implementation. Feedback on the pilot experience was gathered approximately 12 months post-implementation. Clinicians were consistently supportive of the tool’s clinical value and management and support staff reported that the distribution process was stable and caused minimal disruption in patient flow. Qualitative data indicated that the MHPMT did have some clinical utility. For example, one clinician commented, ‘‘It (The MHPMT) allows for a quick summary of symptom status at the beginning of the sessions, it helps to determine the focus of the session.’’ However, there were a number of obstacles to using the MHPMT as a clinical tool. Because
Adm Policy Ment Health
scores were recorded on a third party website and not in the electronic medical record, clinicians often did not check the scores or share these scores with patients’ other medical providers. In addition, frontline clinicians had difficulty understanding the relationship between severity adjusted effect size (a statistical measure of change in a population between 0 and 2), and changes in the MHPMT on an individual basis. There did appear to be marginal improvement in patient satisfaction as the percent excellent ratings in overall satisfaction increased from 34 % for the year before the MHPMT was implemented to 38 % for the year when MHPMT was implemented in the pilot clinic.
MHPMT in the waiting room and brought it with them to the clinician’s office. The clinician did an initial visual scan of the MHPMT and discussed the results with the patient. The clinician either during the visit or at the end of the visit entered each individual item into the EMR. The EMR automatically totaled the items and the EMR also had the capacity to graphically represent current and past MHPMT scores. For the visit note, providers used a documentation template in the EMR which prompted them to review the MHPMT and the last two MHPMT scores were automatically included in the visit note. Providers completed the note either concurrently with the patient or after the visit. Simple job aids were used as part of the training process to explain how to use the tool such as in Fig. 2.
Full Implementation Planning We applied lessons from the pilot experience to inform fullscale implementation. It was important to have a tool that could be easily incorporated in the EMR, was familiar to medical providers, did not involve the administrative complexity of multiple faxing and was less costly. We therefore made the decision to stop contracting with a third-party vendor which included no longer using the vendor developed progress monitoring tool as well as the free standing website. We decided to utilize the Patient Health Questionnaire (PHQ-9) (Kroenke et al. 2001) as the basis for our MHPMT. This measure focuses on depressive symptoms, is also psychometrically valid and had been found to correlate highly with the measure that had been used by the third party vendor. Group Health medical providers were already using the PHQ-9 to track depression and entering it into the EMR. In order to expand the scale to incorporate more nomothetic items to address clinical conditions beyond depression, The Generalized Anxiety Disorder (GAD2), a two-item anxiety scale (Kroenke et al. 2007), a one-item functioning question, the Alcohol Use Disorders Identification Test (Audit C); a screen for alcoholism (Bradley et al. 2007) and a one-item drug question were also added to the scale. These additional questions were entered into the medical EMR as well. Two additional questions related to therapeutic alliance were included in the tool but not entered into the EMR. Figure 1 is a copy of the MHPMT. Changes to how clinicians were trained to use the MHPMT were made based on the pilot trial. The training format was condensed into a 90-min presentation giving a basic overview of the tool, training support staff on how to hand out the tool, and training the clinicians on how to interpret scores and enter them into the electronic medical record. A clinician from the pilot clinic shared their personal experience of the clinical utility of the MHPMT. The process involved support staff distributing the MHPMT at the time of check-into all visits for patients who were 12 years of age or older. The patient completed the
Integration of MHPMT into Clinical Processes of Care While the MHPMT was initially implemented to improve patient satisfaction and track quality of clinical care, it is now being used for a variety of other clinical processes. Question 9 of the PHQ-9 (thoughts of being better off dead or hurting oneself) is used as a standardized way to screen for suicide risk. Questions related to substance use (Audit C) have become part of a department initiative to screen and conduct brief interventions for mental health patients with co-occurring substance use disorders. Clinicians have used the questions related to the therapeutic alliance as a starting point for conversations with patients about issues in the therapeutic relationship, which has been found to be related to patient satisfaction (Crosier et al. 2012). Questions have been added to the MHPMT over time to help address new clinic needs and initiatives. For example, the MHPMT now includes a question on access to firearms as part of clinic efforts to address the patient’s access to lethal means in a standardized manner. A question was also added to assist patients and therapists identify a treatment focus (what are your goals for treatment?) and to promote group therapeutic approaches (Group treatment has been shown to be highly effective in helping people with depression and/or anxiety. Are you interested?). By incorporating the tool into multiple clinical processes, it is kept visible, clinician interest was sustained, and its value was reinforced on both an individual and organizational level.
Organizational Sponsorship and Measurement Targets Organizational change is most likely to be successful when it is in line with business and strategic objectives (Damschroder et al. 2009; Fixsen et al. 2005). Numerous
123
Adm Policy Ment Health Fig. 1 Behavioral health individual progress monitoring tool
Name: __________________________ Date: ___________________________
Behavioral Health Individual Progress Monitoring Tool Over the past two weeks, how often have you been bothered by any of the following problems? 1. 2. 3. 4. 5. 6. 7. 8.
9. 10. 11. 12.
Not at all
More than half the days 2 2 2 2 2
Several Days
Little interest or pleasure in doing things 0 1 Feeling down, depressed or hopeless 0 1 Trouble falling or staying asleep or sleeping too much 0 1 Feeling tired or having little energy 0 1 Poor appetite or overeating 0 1 Feeling bad about yourself or that you are a failure or have 0 1 let yourself or your family down Trouble concentrating on things, such as reading the 0 1 newspaper or watching television Moving or speaking so slowly that other people could have noticed. Or the opposite –being so fidgety or restless that 0 1 you have been moving around a lot Thoughts that you would be better off dead or of hurting 0 1 yourself in some way Feeling nervous, anxious or on edge 0 1 Not being able to stop or control worrying 0 1 Have your problems interfered with your work, family or 0 1 social activities? Please answer these questions about the past four weeks
13. How often do you have a drink containing alcohol? ***If you do not drink, skip to #16 ***
14. How many drinks containing alcohol do you have on a typical day when you are drinking?
3
2
3
2
3
2
3
2 2
3 3
2
3
Never
Monthly or less
2 –4 times a month
2 –3 days per week
≥4 days a week
2
3
4
0
1
1 –2 drinks
3 –4 drinks
5 –6 drinks
7-9 drinks
0
0
1
2
3
0
Less than Monthly 1
0
1
2
Seldom
Fairly often
Never
Never 17. This clinician and I are working on mutually agreed upon goals? 18. This clinician treats me with care and compassion
3 3 3 3 3
2
I do not drink
15. How often do you have 6 or more drinks on one occasion? 16. How often did you use drugs (recreational, marijuana, not prescribed medicines) in the past month?
Nearly every day
10+ drinks 4
Monthly
Weekly
2
3
Daily or almost daily 4
3
4
Very often
Always
0
1
2
3
4
0
1
2
3
4
Rev. 06_19.1_2014
stakeholders, both internal and external, were interested in whether patients involved in mental health treatment were getting better and, in particular, whether recent programmatic initiatives in the department (i.e., care management for chronic psychiatric patients, group therapy for patients with depression and/or anxiety disorders) were leading to enhanced clinical outcomes. Behavioral health leadership was engaged in the adoption of the MHPMT by arguing that a consistently administered progress tool could not only help improve patient satisfaction, but also help objectively track these important quality outcomes. Since the process improvement approach within the organization utilized lean continuous quality improvement principles (Steinfeld et al. 2014), the ability to measure the process was critical to tracking the implementation process. A target was established that progress monitoring tool data would be entered into the EMR for 80 % of all visits for patients 12 years of age and older.
123
Full Implementation Process Once organizational sponsorship was obtained, implementation of the progress monitoring tool became a key improvement process for the Behavioral Health Department. Training as described above was provided along with lean process improvement tools (Steinfeld et al. 2014). These tools included ‘‘standard work’’ (concise descriptions of the process for administration of the MHPMT), visual tracking systems (clinic performance on percentage of adult visits during which the MHPMT was administered) and daily huddles (daily 5–10 min meetings attended by entire team where there was opportunity to discuss tool use and address any process issues). The experience was that within 2 weeks post-training, clinics had fully implemented the progress monitoring tool (i.e., that all patients over the age of 12 were being given the MPHMT at each of their visits and providers were entering the data into the
Adm Policy Ment Health Fig. 2 Progress monitoring tool desktop job aid
Progress Monitoring Tool Desktop Job Aid
Prior to Visit:
•
Look up PHQ-9 flow sheet, graph score or see previous sessions results
Start of Visit:
•
Look at the first 12 questions. If there are mostly 2’s and 3’s, this indicates patient is in some degree of distress (may indicate patient is worse, explore reasons). If mostly 1’s and 2’s, indicates less distress (patient may be doing better, explore what’s going well)
•
Look at suicide question (#9) of PHQ-9. If 2 or 3, do suicide risk assessment.
•
Look at alliance questions (17 and 18). If “always” boxes are checked, acknowledge and validate appreciate that this is working for you). If total less than 8, indicate possible relationship issue and inquire (what’s not working, what could be better).
•
Look at questions 12 –16, if anything greater than zero is checked, inquire about status of substance use.
Results
The MHPMT is administered for approximately 90 % of all visits for patients over the age of 12. This performance has been stable for the past year. This represents approximately 30,000 patient visits a quarter (or almost 120,000 patient visits a year).
Is the Tool Being Used?
Clinician Feedback
In order for a MHPMT to be an effective method for measuring and tracking patient progress, it must be administered on a consistent basis and entered into the EMR. As previously noted, our organization had a goal of entering the MHPMT for 80 % of all visits for patients 12 years of age and older into the EMR. A computer program was developed which reviewed all behavioral health appointments which occurred for patients 12 years of age and older to determine whether a MHPMT was entered into the EMR. Figure 3 summarizes the percentage of visits for patients 12 years of age and older over a 1 year time period in which a MHPMT was administered and entered into the EMR.
An online survey was conducted with staff to provide feedback regarding their experience and satisfaction with the MHPMT. Approximately 50 % of staff responded to the survey. Figure 4 summarizes these results. Staff reported using the MHPMT in a variety of ways. The majority of staff (58 %) responding to this survey reported that they both review past progress scores before patient visits, as well as current progress scores with patients in over 75 % of the visits. Sixty-nine percent of clinicians also reported that they address substance use concerns with a patient if this is identified as an issue (i.e., score greater than zero on substance use question) on the MHPMT in at least half of the visits. Seventy-one percent
EMR). This was in contrast to the 3–6 months implementation experience of the pilot clinic.
123
Adm Policy Ment Health BHS Total
clinical contexts, some clinicians expressed concern that patients may either under-report symptoms (40 % of clinicians reports this occurs in at least 50 % of visits) or overreport symptoms (33 % of clinicians report that this occurs in at least 50 % of the visits) on the MHPMT.
100% 80% 60%
Patient Satisfaction
40% 20% 0%
Fig. 3 Percentage of visits for patients over age of 12 with completed MPHMT in EMR
of clinicians also noted that the MHPMT provides useful information on whether patients are ready for therapy termination. Clinicians were less frequent in their use of the therapeutic alliance questions, with only 38 % of clinicians using this tool in at least 75 % of the visits. While clinicians noted that the MHPMT is useful in a variety of
Fig. 4 Progress monitoring tool survey compilation
In the spring of 2014, Kaiser Permanente Care Management Institute came to Group Health to study how Group Health uses the MHPMT to assess and treat patients with depression. A focus group was conducted with patients to get their feedback on their experience with use of MHPMT. Five middle age patients (4 female, 1 Male) who had a range of treatment experiences with depression (psychiatric, group and individual therapy) from one clinic were identified by their providers as being interested in participating in a feedback session regarding patient experience with the MHPMT. The patients were asked a series of question related to filling out the MHPMT, its value in therapy, its value outside of therapy, how broadly it should be used and patient preferences as it pertains to form completion (in waiting room, paper versus computer
90-100%
75-89%
50-74%
25-49%
What percent of the time do you review the previous session’s progress tool score prior to seeing a patient? (50
44%
14%
2%
14%
What percent of the time do you review with the patient their current progress tool score and explore reasons for the increase or decrease in their stress?(52 responses)
40%
35%
12%
8%
What percent of the time do you discuss with your patient their score on the therapeutic alliance question? (48
13%
25%
8%
15%
What percent of the time do you review and discuss with your patients their substance use when they score above zero on the relevant questions on the progress tool? (53 responses)
23%
31%
15%
15%
What percent of the time do you find the patient’s presentation of their symptoms in session is more severe than what they report on the progress toll (patient under reports on the progress tool?)
0%
16%
24%
33%
What percent of the time do you find the patient’s presentation of their symptoms in session is less severe than what they report on the progress toll (patient over reports on the progress tool?) (51 responses)
0%
16%
17%
37%
Very Helpful
Somewhat Helpful
Neutral
Somewhat Unhelpful
13%
58%
22%
4%
responses)
responses)
How helpful is the progress tool in helping determine when a patient is ready for treatment completion? (52 responses)
123
Adm Policy Ment Health
entries). The session was videotaped and the transcript was reviewed to identify comments that were representative (comments made by eighty percent of patients) of the responses of these five patients to questions asked about their experiences with the MHPMT. Comments included the following: • • • • • •
‘‘It’s almost like starting care before you go into the office.’’ ‘‘It adds structure and focus on what issues I want to talk about during the visit.’’ ‘‘It can be hard to assert yourself. The form helps with the conversation.’’ ‘‘Helps make a ‘‘black cloud’’ of depression concrete.’’ ‘‘Helps me track my depression over time.’’ ‘‘It makes you realize that depression is not a character flaw: It’s a medical condition.’’
Some reservations with the MHPMT were shared with regards to when providers did not review the MHPMT results with patients. Also some of the patients expressed concerns about privacy when filling out the form in the waiting room or data security if the form was filled out on line. The patient feedback speaks to the role of a MHPMT beyond simply a tool to track progress. For the patients in this focus group it was apparent that the tool assisted in understanding their mental health condition, communicating with their provider, and focusing session content. Treatment Outcome A key use of progress monitoring tools is to track individual, clinic, and programmatic outcomes. What would appear to be a simple question (is the patient getting better?) is actually complicated by a number of factors. Patients have episodes of care of different lengths, making it difficult to assign standardized start and end points for treatment. A large proportion of patients complete the PHQ-9 at visits with multiple providers, including their therapist, psychiatrist, and primary care doctor. Do changes in scores have to do with actual changes in the patient’s symptom presentation versus change in setting/therapeutic alliance factors? In addition, treatment goals, including expectable levels of symptom reduction, may be very different in acute versus chronic psychiatric patients. For example, there may be some clinical conditions such as anxiety and/or depression where clinically significant improvement can be expected after time limited treatment (i.e., 6 months or less) time. Yet, there are other clinical conditions such as psychosis, bipolar disorder, borderline personality disorder, where treatment may be more long term (i.e., greater than 6 months) and treatment outcomes may be more focused on symptom stabilization and relapse prevention.
The research literature is not clear in terms of what measure to use and how much change needs to occur in order to determine what is clinically significant change. Based upon consultation from researcher within our organization (Simon 2012) and literature related to treat-totarget (Unutzer et al. 2002), it was decided to break the outcomes tracking into two groups: (a) patients with acute episode of care, which were defined as patients who had been only been seen for 6 months or less and; (b) patients with chronic episode of care, which were defined as patients who had been seen for greater than 6 months of care. Different target parameters were established for each of these subgroups. For the acute care group, the target parameter was percentage of patients who had initial PHQ-9 score of 10 or higher who had 50 % or greater reduction in symptoms comparing the initial PHQ-9 score with the most recent PHQ-9 score (the PHQ-9 score was from any provider who had seen the patient during that 6 month episode of care). Research (Simon 2012; Unutzer et al. 2002) has indicated that the clinically effective range with this measure is between 30 and 60 %. For the chronic patient, outcomes were based upon the most recent PHQ-9 score and divided into three categories • Unimproved (PHQ-9 score greater than 10). • Partial remission (PHQ-9 score in between 5 and 10). • Remission (PHQ-9 score less than 5). The reason for breaking the population of patients into two different groups is that it was hypothesized that there are a group of patients who have chronic psychiatric conditions who are likely going to be in care for extended periods of time (i.e., greater than 6 months) where 50 % improvement in symptoms may not be a realistic target. In addition, these patients may be stable (i.e., current symptoms below clinical threshold) but ongoing treatment may be necessary to maintain symptom stability and prevent relapse. Data was collected at the individual, clinic, and department level. Figure 5 shows the results at the department level for acute care patients. Department level performance is consistently at the 40 % effectiveness range, meaning that approximately 40 % of the patients have experienced at least a 50 % reduction in PHQ-9 scores. The performance is consistent across clinics, consistent across time, and consistent with the literature (Simon 2012; Unutzer et al. 2002). There has also been some analysis of the data on the programmatic level. For example, to compare the effectiveness of a new group therapy program to individual therapy, outcome data has indicated equivalencies in the efficacy for patients in individual as compared to group therapy. Thirty-nine percent of patients who utilized group therapy for treatment of depression and/or anxiety experienced a 50 % reduction in
123
Adm Policy Ment Health Department acute Care outcomes
Department
100%
100% 80%
80%
60% 40%
2240
3508
3700
4377
4652
5035
20%
Percent of total patients
Q2 2012
Q3 2012
Q4 2012
Q1 2013
Q2 2013
Q3 2013
123
34.5%
38.4%
32.4%
31.8%
33.1%
29.8%
40%
14645
13834
14410
14729
32.8%
32.7%
31.0%
31.0%
34.1%
34.2%
33.6%
33.4%
33.1%
33.2%
35.4%
35.5%
0% Q2 2012 Q3 2012 Q4 2012 Q1 2013 Q2 2013 Q3 2013
Fig. 5 Percentage of patient’s in acute care with 50 % reduction in symptoms
symptoms, as compared to 40 % of patients in individual therapy who have experienced a 50 % reduction in symptoms. In Fig. 6, the graph depicts patients during five calendar quarters that are currently in treatment and have had a treatment episode of 6 months or longer. Each quarter does represent some overlap of patients as some patients continued in treatment from one calendar quarter to the next. The top part of each bar graph identifies the percentage of patients whose last PHQ-9 score was greater than 10. The middle part of each bar graph includes the percentage of patients whose last PHQ-9 score was 5–10 and the bottom part of each bar graph identifies the percentage of patients whose last PHQ-9 score was less than 5. The cutoffs for partial remission and remission on the PHQ-9 are consistent with literature looking at relationship between scores on the PHQ-9 and depressive symptoms (Unutzer et al. 2002). As it pertains to patients in chronic care, there appears to be equal percentage in the unimproved, partial remission, and full remission categories. Targets for this chronic care group have not been established. What appears to be of most interest is that 30 % of these patients appear to be in full remission (i.e., PHQ-9 score less than 5). It is not clear whether these patients continue to need episodic care or are stable and considered to be treatment completers. In addition, we have been hesitant to share providerspecific performance data as it pertains to outcomes. We are concerned about the impact sharing such data would have on the engagement of low-performing staff, and we cannot be sure that variations in scores are not a result of other factors not related to individual provider performance. What we have done is to develop a report that is more descriptive in nature which describes a providers case load (i.e., how many patients they are seeing, how long they have seen these patients for, what other disciplines are involved in their care) as well as information regarding changes in their MHPMT score (how many of their patients most recent PHQ-9 score is above 10 or below 5). We have begun to share this report with clinic managers and staff with the intent of facilitating a better understanding of
13985
60%
20%
0%
14785
Severity Categories Unimproved (PHQ9 of >10) Partial Remission (PHQ9 of 5 - 10) Remission (PHQ9 of
ORIGINAL PAPER
Progress Monitoring in an Integrated Health Care System: Tracking Behavioral Health Vital Signs Bradley Steinfeld1 • Allie Franklin2 • Brian Mercer1 • Rebecca Fraynt1 Greg Simon3
•
Springer Science+Business Media New York 2015
Abstract Progress monitoring implementation in an integrated health care system is a complex process that must address factors such as measurement, technology, delivery system care processes, patient needs and provider requirements. This article will describe how one organization faced these challenges by identifying the key decision points (choice of measure, process for completing rating scale, interface with electronic medical record and clinician engagement) critical to implementation. Qualitative and quantitative data will be presented describing customer and stakeholder satisfaction with the mental health progress monitoring tool (MHPMT) as well as organizational performance with key measurement targets. Keywords Progress monitoring Behavioral health Integrated health care system
Introduction It is well documented that measurement of performance and providing feedback, often referred to as progress monitoring (Bickman et al. 2011; Lambert 2010), is foundational to improvement in mental health care. Institute of Medicine of the National Academies (2001, 2006) recommended that both mental health and health care providers regularly use valid and reliable patient & Bradley Steinfeld [email protected] 1
Group Health Cooperative, 950 Pacific Ave Suite 900, Tacoma, WA 98402, USA
2
Optum Health Care, Tacoma, USA
3
Group Health Research Institute, Seattle, USA
assessment instruments to measure treatment progress and outcomes. Despite the research evidence that progress monitoring improves treatment outcomes (Lambert 2010), as well as the development of a number of mental health progress monitoring systems (Bickman et al. 2011; Chorpita and Daleiden 2010; Duncan et al. 2010; Lambert 2010), the majority of clinicians do not use these tools as a routine part of their practice. Many of the barriers to implementation are similar to those identified in implementation of evidence based practices (EBP). These include lack of training, staff engagement, organizational readiness and implementation structures (Fixsen et al. 2005; McHugh and Barlow 2012; Steinfeld et al. 2014). In addition, there are a number of barriers as it pertains to clinician attitudes including time and cost to use tool (Jensen-Doss and Hawley 2010), clinician hesitance to adopt computer-based technology (Lambert 2010), clinician belief they are in a better position to judge patient progress than the patients themselves and process barriers such as administrative complexity and cost (Lambert 2010). As behavioral health care is increasingly integrated within health care delivery systems (McDaniel and Degruy 2014), there are additional implementation challenges. Current behavioral health progress monitoring systems are typically only familiar to mental health practitioners and are tracked using stand-alone platforms that do not integrate easily with fully integrated electronic medical records or chronic disease management programs (Steinfeld and Keyes 2011). This article will describe the experience of implementing a Mental Health Progress Monitoring Tool (MHPMT) in an integrated health care delivery system. An overview of the setting in which this work occurred will be provided along with the contextual factors which led this
123
Adm Policy Ment Health
organization to develop a MHPMT system. The process of implementation will be described including the measurement selection, administrative, and training processes, as well as how pilot implementation influenced full-scale implementation. Full-scale implementation will be described with a focus on staff engagement, integration of MHPMT with electronic medical record (EMR), organizational sponsorship, and measurement targets. Results will be shared which will include patient and staff qualitative feedback, rates of MHPMT usage and patient outcomes. Finally, critical factors leading to successful implementation, as well as continued challenges, will be shared.
The Setting Group Health Cooperative is a consumer-governed, nonprofit, health care system that provides coverage and heath care for approximately 650,000 people in Washington and Idaho. Medical care is provided in 26 medical center and clinics by over 1000 medical providers. The Behavioral Health Service Department at Group Health Cooperative provides coverage, mental health and chemical dependency care for 17,000 people per year via its seven model mental health clinics, as well as its contracted network. The target audience implementing the progress monitoring tool consisted of the mental health providers in these seven staff model clinics (approximately 125 staff), composed of master level clinicians, psychologists, psychiatrists and psychiatric nurses.
Getting Started/A Case for Change For a number of years, Group Health had been focused on improving patient satisfaction as measured by patient response to surveys assessing satisfaction with overall treatment (Crosier et al. 2012). Some initial work indicated that regular use of progress monitoring tools such as the Burns Mood Enhancement Scale (Burns 1997) and Outcome Rating Scale (Miller et al. 2003), increased patient satisfaction scores. However, this initiative was not continued as there were a number of barriers with these progress monitoring tools, which included clinician dissatisfaction with items on the scales, difficulty tracking data over time, and difficulty consistently administering the scales to patients.
Pilot Program Subsequently, one of the mental health clinics had low patient satisfaction survey scores, and was interested in again piloting whether systematic use of a progress monitoring tool
123
could improve patient satisfaction scores. This pilot program involved one group health behavioral health clinic made up of 13 clinicians (2 psychiatrists, 2 psychiatric nurses, 2 psychologists, and 7 master level therapists) all of whom participated in the progress monitoring pilot. In order to address the difficulties encountered during the initial, smallscale study, a third party vendor was solicited to implement a MHPMT. A key issue was deciding what items should be on the progress monitoring measure. As Scott and Lewis (2015) have noted, an important consideration is determining whether the items should be idiographic (i.e., tailored to the patient’s specific problem) or nomothetic (i.e., standardized self-report measures that can be administered consistently over time). Based upon our initial experience and in consultation with the third party vendor, we decided to use a nomothetic approach which would simplify administration (same measure for each patient at each visit) and we used items that the vendor had identified were sensitive to change in a wide variety of clinical conditions (i.e., global distress) and were also psychometrically valid. Severity adjusted effect size, a standardized method for determining change that takes into account case mix and symptoms severity (Brown and Minami 2010) was the measure used by the vendor to determine clinical significance of change across a patient population. The vendor also had a free standing web site which included raw scores on an individual basis, a graphical representation of the change in raw scores for individual patients, and severity effect change scores for population of patients on the provider, clinic, and organizational level. To assure that patients consistently completed the MHPMT, support staff distributed the forms to all adult patients over the age of 12 at all visits. Patients gave the completed forms to their providers, who then returned the forms to the front desk staff. These forms were then faxed to the third-party vendor on a daily basis. Clinicians were also trained to use the MHPMT through didactic presentations. Information was provided on theoretical/clinical rationale for progress monitoring scales, how to interpret the scales and discuss the results with patients, and how to access the website of the third party vendor. Follow-up consultations occurred on a monthly basis for 2–3 months post-implementation. Feedback on the pilot experience was gathered approximately 12 months post-implementation. Clinicians were consistently supportive of the tool’s clinical value and management and support staff reported that the distribution process was stable and caused minimal disruption in patient flow. Qualitative data indicated that the MHPMT did have some clinical utility. For example, one clinician commented, ‘‘It (The MHPMT) allows for a quick summary of symptom status at the beginning of the sessions, it helps to determine the focus of the session.’’ However, there were a number of obstacles to using the MHPMT as a clinical tool. Because
Adm Policy Ment Health
scores were recorded on a third party website and not in the electronic medical record, clinicians often did not check the scores or share these scores with patients’ other medical providers. In addition, frontline clinicians had difficulty understanding the relationship between severity adjusted effect size (a statistical measure of change in a population between 0 and 2), and changes in the MHPMT on an individual basis. There did appear to be marginal improvement in patient satisfaction as the percent excellent ratings in overall satisfaction increased from 34 % for the year before the MHPMT was implemented to 38 % for the year when MHPMT was implemented in the pilot clinic.
MHPMT in the waiting room and brought it with them to the clinician’s office. The clinician did an initial visual scan of the MHPMT and discussed the results with the patient. The clinician either during the visit or at the end of the visit entered each individual item into the EMR. The EMR automatically totaled the items and the EMR also had the capacity to graphically represent current and past MHPMT scores. For the visit note, providers used a documentation template in the EMR which prompted them to review the MHPMT and the last two MHPMT scores were automatically included in the visit note. Providers completed the note either concurrently with the patient or after the visit. Simple job aids were used as part of the training process to explain how to use the tool such as in Fig. 2.
Full Implementation Planning We applied lessons from the pilot experience to inform fullscale implementation. It was important to have a tool that could be easily incorporated in the EMR, was familiar to medical providers, did not involve the administrative complexity of multiple faxing and was less costly. We therefore made the decision to stop contracting with a third-party vendor which included no longer using the vendor developed progress monitoring tool as well as the free standing website. We decided to utilize the Patient Health Questionnaire (PHQ-9) (Kroenke et al. 2001) as the basis for our MHPMT. This measure focuses on depressive symptoms, is also psychometrically valid and had been found to correlate highly with the measure that had been used by the third party vendor. Group Health medical providers were already using the PHQ-9 to track depression and entering it into the EMR. In order to expand the scale to incorporate more nomothetic items to address clinical conditions beyond depression, The Generalized Anxiety Disorder (GAD2), a two-item anxiety scale (Kroenke et al. 2007), a one-item functioning question, the Alcohol Use Disorders Identification Test (Audit C); a screen for alcoholism (Bradley et al. 2007) and a one-item drug question were also added to the scale. These additional questions were entered into the medical EMR as well. Two additional questions related to therapeutic alliance were included in the tool but not entered into the EMR. Figure 1 is a copy of the MHPMT. Changes to how clinicians were trained to use the MHPMT were made based on the pilot trial. The training format was condensed into a 90-min presentation giving a basic overview of the tool, training support staff on how to hand out the tool, and training the clinicians on how to interpret scores and enter them into the electronic medical record. A clinician from the pilot clinic shared their personal experience of the clinical utility of the MHPMT. The process involved support staff distributing the MHPMT at the time of check-into all visits for patients who were 12 years of age or older. The patient completed the
Integration of MHPMT into Clinical Processes of Care While the MHPMT was initially implemented to improve patient satisfaction and track quality of clinical care, it is now being used for a variety of other clinical processes. Question 9 of the PHQ-9 (thoughts of being better off dead or hurting oneself) is used as a standardized way to screen for suicide risk. Questions related to substance use (Audit C) have become part of a department initiative to screen and conduct brief interventions for mental health patients with co-occurring substance use disorders. Clinicians have used the questions related to the therapeutic alliance as a starting point for conversations with patients about issues in the therapeutic relationship, which has been found to be related to patient satisfaction (Crosier et al. 2012). Questions have been added to the MHPMT over time to help address new clinic needs and initiatives. For example, the MHPMT now includes a question on access to firearms as part of clinic efforts to address the patient’s access to lethal means in a standardized manner. A question was also added to assist patients and therapists identify a treatment focus (what are your goals for treatment?) and to promote group therapeutic approaches (Group treatment has been shown to be highly effective in helping people with depression and/or anxiety. Are you interested?). By incorporating the tool into multiple clinical processes, it is kept visible, clinician interest was sustained, and its value was reinforced on both an individual and organizational level.
Organizational Sponsorship and Measurement Targets Organizational change is most likely to be successful when it is in line with business and strategic objectives (Damschroder et al. 2009; Fixsen et al. 2005). Numerous
123
Adm Policy Ment Health Fig. 1 Behavioral health individual progress monitoring tool
Name: __________________________ Date: ___________________________
Behavioral Health Individual Progress Monitoring Tool Over the past two weeks, how often have you been bothered by any of the following problems? 1. 2. 3. 4. 5. 6. 7. 8.
9. 10. 11. 12.
Not at all
More than half the days 2 2 2 2 2
Several Days
Little interest or pleasure in doing things 0 1 Feeling down, depressed or hopeless 0 1 Trouble falling or staying asleep or sleeping too much 0 1 Feeling tired or having little energy 0 1 Poor appetite or overeating 0 1 Feeling bad about yourself or that you are a failure or have 0 1 let yourself or your family down Trouble concentrating on things, such as reading the 0 1 newspaper or watching television Moving or speaking so slowly that other people could have noticed. Or the opposite –being so fidgety or restless that 0 1 you have been moving around a lot Thoughts that you would be better off dead or of hurting 0 1 yourself in some way Feeling nervous, anxious or on edge 0 1 Not being able to stop or control worrying 0 1 Have your problems interfered with your work, family or 0 1 social activities? Please answer these questions about the past four weeks
13. How often do you have a drink containing alcohol? ***If you do not drink, skip to #16 ***
14. How many drinks containing alcohol do you have on a typical day when you are drinking?
3
2
3
2
3
2
3
2 2
3 3
2
3
Never
Monthly or less
2 –4 times a month
2 –3 days per week
≥4 days a week
2
3
4
0
1
1 –2 drinks
3 –4 drinks
5 –6 drinks
7-9 drinks
0
0
1
2
3
0
Less than Monthly 1
0
1
2
Seldom
Fairly often
Never
Never 17. This clinician and I are working on mutually agreed upon goals? 18. This clinician treats me with care and compassion
3 3 3 3 3
2
I do not drink
15. How often do you have 6 or more drinks on one occasion? 16. How often did you use drugs (recreational, marijuana, not prescribed medicines) in the past month?
Nearly every day
10+ drinks 4
Monthly
Weekly
2
3
Daily or almost daily 4
3
4
Very often
Always
0
1
2
3
4
0
1
2
3
4
Rev. 06_19.1_2014
stakeholders, both internal and external, were interested in whether patients involved in mental health treatment were getting better and, in particular, whether recent programmatic initiatives in the department (i.e., care management for chronic psychiatric patients, group therapy for patients with depression and/or anxiety disorders) were leading to enhanced clinical outcomes. Behavioral health leadership was engaged in the adoption of the MHPMT by arguing that a consistently administered progress tool could not only help improve patient satisfaction, but also help objectively track these important quality outcomes. Since the process improvement approach within the organization utilized lean continuous quality improvement principles (Steinfeld et al. 2014), the ability to measure the process was critical to tracking the implementation process. A target was established that progress monitoring tool data would be entered into the EMR for 80 % of all visits for patients 12 years of age and older.
123
Full Implementation Process Once organizational sponsorship was obtained, implementation of the progress monitoring tool became a key improvement process for the Behavioral Health Department. Training as described above was provided along with lean process improvement tools (Steinfeld et al. 2014). These tools included ‘‘standard work’’ (concise descriptions of the process for administration of the MHPMT), visual tracking systems (clinic performance on percentage of adult visits during which the MHPMT was administered) and daily huddles (daily 5–10 min meetings attended by entire team where there was opportunity to discuss tool use and address any process issues). The experience was that within 2 weeks post-training, clinics had fully implemented the progress monitoring tool (i.e., that all patients over the age of 12 were being given the MPHMT at each of their visits and providers were entering the data into the
Adm Policy Ment Health Fig. 2 Progress monitoring tool desktop job aid
Progress Monitoring Tool Desktop Job Aid
Prior to Visit:
•
Look up PHQ-9 flow sheet, graph score or see previous sessions results
Start of Visit:
•
Look at the first 12 questions. If there are mostly 2’s and 3’s, this indicates patient is in some degree of distress (may indicate patient is worse, explore reasons). If mostly 1’s and 2’s, indicates less distress (patient may be doing better, explore what’s going well)
•
Look at suicide question (#9) of PHQ-9. If 2 or 3, do suicide risk assessment.
•
Look at alliance questions (17 and 18). If “always” boxes are checked, acknowledge and validate appreciate that this is working for you). If total less than 8, indicate possible relationship issue and inquire (what’s not working, what could be better).
•
Look at questions 12 –16, if anything greater than zero is checked, inquire about status of substance use.
Results
The MHPMT is administered for approximately 90 % of all visits for patients over the age of 12. This performance has been stable for the past year. This represents approximately 30,000 patient visits a quarter (or almost 120,000 patient visits a year).
Is the Tool Being Used?
Clinician Feedback
In order for a MHPMT to be an effective method for measuring and tracking patient progress, it must be administered on a consistent basis and entered into the EMR. As previously noted, our organization had a goal of entering the MHPMT for 80 % of all visits for patients 12 years of age and older into the EMR. A computer program was developed which reviewed all behavioral health appointments which occurred for patients 12 years of age and older to determine whether a MHPMT was entered into the EMR. Figure 3 summarizes the percentage of visits for patients 12 years of age and older over a 1 year time period in which a MHPMT was administered and entered into the EMR.
An online survey was conducted with staff to provide feedback regarding their experience and satisfaction with the MHPMT. Approximately 50 % of staff responded to the survey. Figure 4 summarizes these results. Staff reported using the MHPMT in a variety of ways. The majority of staff (58 %) responding to this survey reported that they both review past progress scores before patient visits, as well as current progress scores with patients in over 75 % of the visits. Sixty-nine percent of clinicians also reported that they address substance use concerns with a patient if this is identified as an issue (i.e., score greater than zero on substance use question) on the MHPMT in at least half of the visits. Seventy-one percent
EMR). This was in contrast to the 3–6 months implementation experience of the pilot clinic.
123
Adm Policy Ment Health BHS Total
clinical contexts, some clinicians expressed concern that patients may either under-report symptoms (40 % of clinicians reports this occurs in at least 50 % of visits) or overreport symptoms (33 % of clinicians report that this occurs in at least 50 % of the visits) on the MHPMT.
100% 80% 60%
Patient Satisfaction
40% 20% 0%
Fig. 3 Percentage of visits for patients over age of 12 with completed MPHMT in EMR
of clinicians also noted that the MHPMT provides useful information on whether patients are ready for therapy termination. Clinicians were less frequent in their use of the therapeutic alliance questions, with only 38 % of clinicians using this tool in at least 75 % of the visits. While clinicians noted that the MHPMT is useful in a variety of
Fig. 4 Progress monitoring tool survey compilation
In the spring of 2014, Kaiser Permanente Care Management Institute came to Group Health to study how Group Health uses the MHPMT to assess and treat patients with depression. A focus group was conducted with patients to get their feedback on their experience with use of MHPMT. Five middle age patients (4 female, 1 Male) who had a range of treatment experiences with depression (psychiatric, group and individual therapy) from one clinic were identified by their providers as being interested in participating in a feedback session regarding patient experience with the MHPMT. The patients were asked a series of question related to filling out the MHPMT, its value in therapy, its value outside of therapy, how broadly it should be used and patient preferences as it pertains to form completion (in waiting room, paper versus computer
90-100%
75-89%
50-74%
25-49%
What percent of the time do you review the previous session’s progress tool score prior to seeing a patient? (50
44%
14%
2%
14%
What percent of the time do you review with the patient their current progress tool score and explore reasons for the increase or decrease in their stress?(52 responses)
40%
35%
12%
8%
What percent of the time do you discuss with your patient their score on the therapeutic alliance question? (48
13%
25%
8%
15%
What percent of the time do you review and discuss with your patients their substance use when they score above zero on the relevant questions on the progress tool? (53 responses)
23%
31%
15%
15%
What percent of the time do you find the patient’s presentation of their symptoms in session is more severe than what they report on the progress toll (patient under reports on the progress tool?)
0%
16%
24%
33%
What percent of the time do you find the patient’s presentation of their symptoms in session is less severe than what they report on the progress toll (patient over reports on the progress tool?) (51 responses)
0%
16%
17%
37%
Very Helpful
Somewhat Helpful
Neutral
Somewhat Unhelpful
13%
58%
22%
4%
responses)
responses)
How helpful is the progress tool in helping determine when a patient is ready for treatment completion? (52 responses)
123
Adm Policy Ment Health
entries). The session was videotaped and the transcript was reviewed to identify comments that were representative (comments made by eighty percent of patients) of the responses of these five patients to questions asked about their experiences with the MHPMT. Comments included the following: • • • • • •
‘‘It’s almost like starting care before you go into the office.’’ ‘‘It adds structure and focus on what issues I want to talk about during the visit.’’ ‘‘It can be hard to assert yourself. The form helps with the conversation.’’ ‘‘Helps make a ‘‘black cloud’’ of depression concrete.’’ ‘‘Helps me track my depression over time.’’ ‘‘It makes you realize that depression is not a character flaw: It’s a medical condition.’’
Some reservations with the MHPMT were shared with regards to when providers did not review the MHPMT results with patients. Also some of the patients expressed concerns about privacy when filling out the form in the waiting room or data security if the form was filled out on line. The patient feedback speaks to the role of a MHPMT beyond simply a tool to track progress. For the patients in this focus group it was apparent that the tool assisted in understanding their mental health condition, communicating with their provider, and focusing session content. Treatment Outcome A key use of progress monitoring tools is to track individual, clinic, and programmatic outcomes. What would appear to be a simple question (is the patient getting better?) is actually complicated by a number of factors. Patients have episodes of care of different lengths, making it difficult to assign standardized start and end points for treatment. A large proportion of patients complete the PHQ-9 at visits with multiple providers, including their therapist, psychiatrist, and primary care doctor. Do changes in scores have to do with actual changes in the patient’s symptom presentation versus change in setting/therapeutic alliance factors? In addition, treatment goals, including expectable levels of symptom reduction, may be very different in acute versus chronic psychiatric patients. For example, there may be some clinical conditions such as anxiety and/or depression where clinically significant improvement can be expected after time limited treatment (i.e., 6 months or less) time. Yet, there are other clinical conditions such as psychosis, bipolar disorder, borderline personality disorder, where treatment may be more long term (i.e., greater than 6 months) and treatment outcomes may be more focused on symptom stabilization and relapse prevention.
The research literature is not clear in terms of what measure to use and how much change needs to occur in order to determine what is clinically significant change. Based upon consultation from researcher within our organization (Simon 2012) and literature related to treat-totarget (Unutzer et al. 2002), it was decided to break the outcomes tracking into two groups: (a) patients with acute episode of care, which were defined as patients who had been only been seen for 6 months or less and; (b) patients with chronic episode of care, which were defined as patients who had been seen for greater than 6 months of care. Different target parameters were established for each of these subgroups. For the acute care group, the target parameter was percentage of patients who had initial PHQ-9 score of 10 or higher who had 50 % or greater reduction in symptoms comparing the initial PHQ-9 score with the most recent PHQ-9 score (the PHQ-9 score was from any provider who had seen the patient during that 6 month episode of care). Research (Simon 2012; Unutzer et al. 2002) has indicated that the clinically effective range with this measure is between 30 and 60 %. For the chronic patient, outcomes were based upon the most recent PHQ-9 score and divided into three categories • Unimproved (PHQ-9 score greater than 10). • Partial remission (PHQ-9 score in between 5 and 10). • Remission (PHQ-9 score less than 5). The reason for breaking the population of patients into two different groups is that it was hypothesized that there are a group of patients who have chronic psychiatric conditions who are likely going to be in care for extended periods of time (i.e., greater than 6 months) where 50 % improvement in symptoms may not be a realistic target. In addition, these patients may be stable (i.e., current symptoms below clinical threshold) but ongoing treatment may be necessary to maintain symptom stability and prevent relapse. Data was collected at the individual, clinic, and department level. Figure 5 shows the results at the department level for acute care patients. Department level performance is consistently at the 40 % effectiveness range, meaning that approximately 40 % of the patients have experienced at least a 50 % reduction in PHQ-9 scores. The performance is consistent across clinics, consistent across time, and consistent with the literature (Simon 2012; Unutzer et al. 2002). There has also been some analysis of the data on the programmatic level. For example, to compare the effectiveness of a new group therapy program to individual therapy, outcome data has indicated equivalencies in the efficacy for patients in individual as compared to group therapy. Thirty-nine percent of patients who utilized group therapy for treatment of depression and/or anxiety experienced a 50 % reduction in
123
Adm Policy Ment Health Department acute Care outcomes
Department
100%
100% 80%
80%
60% 40%
2240
3508
3700
4377
4652
5035
20%
Percent of total patients
Q2 2012
Q3 2012
Q4 2012
Q1 2013
Q2 2013
Q3 2013
123
34.5%
38.4%
32.4%
31.8%
33.1%
29.8%
40%
14645
13834
14410
14729
32.8%
32.7%
31.0%
31.0%
34.1%
34.2%
33.6%
33.4%
33.1%
33.2%
35.4%
35.5%
0% Q2 2012 Q3 2012 Q4 2012 Q1 2013 Q2 2013 Q3 2013
Fig. 5 Percentage of patient’s in acute care with 50 % reduction in symptoms
symptoms, as compared to 40 % of patients in individual therapy who have experienced a 50 % reduction in symptoms. In Fig. 6, the graph depicts patients during five calendar quarters that are currently in treatment and have had a treatment episode of 6 months or longer. Each quarter does represent some overlap of patients as some patients continued in treatment from one calendar quarter to the next. The top part of each bar graph identifies the percentage of patients whose last PHQ-9 score was greater than 10. The middle part of each bar graph includes the percentage of patients whose last PHQ-9 score was 5–10 and the bottom part of each bar graph identifies the percentage of patients whose last PHQ-9 score was less than 5. The cutoffs for partial remission and remission on the PHQ-9 are consistent with literature looking at relationship between scores on the PHQ-9 and depressive symptoms (Unutzer et al. 2002). As it pertains to patients in chronic care, there appears to be equal percentage in the unimproved, partial remission, and full remission categories. Targets for this chronic care group have not been established. What appears to be of most interest is that 30 % of these patients appear to be in full remission (i.e., PHQ-9 score less than 5). It is not clear whether these patients continue to need episodic care or are stable and considered to be treatment completers. In addition, we have been hesitant to share providerspecific performance data as it pertains to outcomes. We are concerned about the impact sharing such data would have on the engagement of low-performing staff, and we cannot be sure that variations in scores are not a result of other factors not related to individual provider performance. What we have done is to develop a report that is more descriptive in nature which describes a providers case load (i.e., how many patients they are seeing, how long they have seen these patients for, what other disciplines are involved in their care) as well as information regarding changes in their MHPMT score (how many of their patients most recent PHQ-9 score is above 10 or below 5). We have begun to share this report with clinic managers and staff with the intent of facilitating a better understanding of
13985
60%
20%
0%
14785
Severity Categories Unimproved (PHQ9 of >10) Partial Remission (PHQ9 of 5 - 10) Remission (PHQ9 of
