Abstract
PURPOSE
Guideline implementation in primary care has proven difficult. Although external assistance through performance feedback, academic detailing, practice facilitation (PF), and learning collaboratives seems to help, the best combination of interventions has not been determined.
METHODS
In a cluster randomized trial, we compared the independent and combined effectiveness of PF and local learning collaboratives (LLCs), combined with performance feedback and academic detailing, with performance feedback and academic detailing alone on implementation of the National Heart, Lung and Blood Institute’s Asthma Guidelines. The study was conducted in 3 primary care practice-based research networks. Medical records of patients with asthma seen during pre- and postintervention periods were abstracted to determine adherence to 6 guideline recommendations. McNemar’s test and multivariate modeling were used to evaluate the impact of the interventions.
RESULTS
Across 43 practices, 1,016 patients met inclusion criteria. Overall, adherence to all 6 recommendations increased (P ≤.002). Examination of improvement by study arm in unadjusted analyses showed that practices in the control arm significantly improved adherence to 2 of 6 recommendations, whereas practices in the PF arm improved in 3, practices in the LLCs improved in 4, and practices in the PF + LLC arm improved in 5 of 6 recommendations. In multivariate modeling, PF practices significantly improved assessment of asthma severity (odds ratio [OR] = 2.5, 95% CI, 1.7–3.8) and assessment of asthma level of control (OR = 2.3, 95% CI, 1.5–3.5) compared with control practices. Practices assigned to LLCs did not improve significantly more than control practices for any recommendation.
CONCLUSIONS
Addition of PF to performance feedback and academic detailing was helpful to practices attempting to improve adherence to asthma guidelines.
Keywords: asthma, primary health care, practice-based research network, practice guidelines as topic, quality improvement, practice facilitation
INTRODUCTION
Implementation of clinical practice guidelines for managing chronic diseases can be challenging for primary care practices. Despite the availability since 1991 of the guidelines sponsored by the National Heart, Lung and Blood Institute (NHLBI) and produced by the National Asthma Education and Prevention Program (NAEPP), fewer than 40% of asthma care notes include information about symptoms, missed activities, and environmental triggers.1–4 A 2001 survey of more than 60,000 patients treated by 4,901 primary care physicians found that while 62% of pediatric patients and 68% of adult patients reported more than 2 symptomatic days in the past week, only 60% had a prescription for a controller medication.5 Other studies found that practices prescribing a corticosteroid inhaler ranged from 15% among patients with moderate to severe asthma to 72% among patients with severe asthma.6,7 Between 0% and 50% of parents of asthmatic children report having received an action plan.7–10
Simple dissemination strategies, combined with clinician education, toolkits, and performance feedback, have little impact on guideline implementation.11–15 More intensive forms of assistance and multicomponent interventions may be more helpful.16–19 Two promising types of assistance are practice facilitation (PF)20–22 and learning collaboratives.23–27 Assumptions behind PF are that many practices are inadequately resourced, lack the experience and skills required, and are so unique that each must implement innovations differently. Relationships between practice facilitators and practice staff appear to be critical to success, resembling the Cooperative Extension, where agents develop relationships with families to facilitate implementation of evidence-based farming practices.28
Peer-to-peer learning has been found to motivate changes in practice.29 Learning collaboratives create opportunities for improvement within a framework of competition and urgency. Learning collaboratives typically involve large numbers of practices that receive education, perform medical record reviews, develop registries, and work collaboratively to implement evidence-based strategies. Although studies evaluating the effectiveness of collaboratives have produced mixed results, collaboratives remain a widely used approach.30 Our experience suggests that the same motivation, competition, and collaborative learning generated in large collaboratives can be achieved by small numbers of practices meeting more often and for shorter periods of time, and that these local learning collaboratives (LLCs) are well-accepted by clinicians and staff.31
There is no literature directly comparing the effectiveness of PF and LLCs. We therefore conducted a cluster randomized trial comparing the 2 interventions, separately and together, with performance feedback and academic detailing alone on implementation of asthma guideline recommendations.
METHODS
Study Design
The study was a 6-month, cluster randomized controlled trial, in which we randomly assigned practices in close geographic proximity (to facilitate LLC implementation) in clusters of 3 to 1 of 4 study arms: (1) PF alone, (2) LLC alone, (3) PF + LLC, and (4) control. We focused on 6 key guideline recommendations from the NAEPP,32 based upon input from our asthma experts (K.E., B.Y.): (1) documentation of intermittent vs persistent asthma (severity assessment), (2) assessment of exposure to environmental triggers, (3) assessment of level of control, (4) prescription of controller medications for persistent asthma, (5) a written asthma action plan, and (6) planned asthma visits. Participating primary care practices were members of 3 practice-based research networks, 1 in Oklahoma and 2 in western New York.
Interventions
All practices received performance feedback, academic detailing, summaries of NHLBI asthma guidelines, and a toolkit, which included the Asthma Control Test, the Asthma APGAR,33 and action plan templates in English and Spanish. A clinician to help each practice develop an improvement plan provided academic detailing, including performance feedback based on record abstractions and information on best indigenous practices.
In addition, practices assigned to PF and PF + LLC received assistance from a practice facilitator who visited practices for a half-day weekly or a full day every other week for 6 months to assist the practice in meeting their target goals.
Practices assigned to LLC and PF + LLC were expected to meet once a month for at least 1 hour for 6 months to review each other’s performance data, discuss successful and unsuccessful strategies, and refine plans for implementing guideline recommendations.
Data Collection
We used PF field notes and LLC minutes to assess fidelity to intervention expectations. Eighteen months after the start of the intervention, patient records were abstracted by trained medical record abstractors using standardized forms to collect information pertinent to the 6 guideline recommendations. For each practice, the goal was to abstract information from 60 records of patients meeting the following criteria: an International Classification of Disease, Ninth Revision (ICD-9) code for asthma (493.xxx), born between 1956 and 2004, and at least 1 visit for asthma between 9 months before and 18 months after the academic detailing session. Reasons for visits were classified as primary asthma-related visit (including routine follow-up for asthma and symptomatic asthma), other asthma-related visit (visit for another concern where asthma was addressed), and non–asthma-related visit (vaccinations, other acute or chronic disease visits where asthma was not addressed). We neither abstracted nor counted non–asthma-related visits during the process.
Key Outcomes
Evidence of guideline implementation was determined during 2 periods: 9-months preintervention and 18-months postintervention (6 months of active intervention and 12-months afterward). Information from all asthma-related visits during each period was used in determining evidence for implementation of each guideline recommendation.
We considered asthma severity to be assessed if the type of asthma was ever described as intermittent, mild, moderate, or severe persistent, exercise-induced bronchospasm, or asthmatic bronchitis, or if spirometry showed a forced expiratory volume in the first second of expiration (FEV1) of less than 80%. We considered any mention of asthma triggers or evidence that patients were asked about exposure to secondhand smoke as assessment of environmental triggers. An asthma action plan had to be evident in the medical record. The denominator for asthma severity, assessment of environmental triggers, and asthma action plan was all active patients with at least 1 asthma-related visit in both time periods.
Assessment of asthma level of control was defined as documentation of at least 3 of the following: daytime symptoms, nighttime symptoms, symptoms while exercising (or playing), missing school or work, use of rescue inhaler, and office measurement of peak expiratory flow rate. The denominator was the total number of asthma-related visits in each period.
We looked for evidence of a prescription for asthma controller medication (inhaled corticosteroid, inhaled corticosteroid plus a long-acting beta agonist, or leukotriene inhibitor) among persistent asthmatic patients, defined as all asthmatic patients referred to as having persistent asthma or given a controller medication. The denominator was all patients with persistent asthma seen at least once during both periods. We examined the number of patients whose clinician recommended a follow-up visit in the next 12 months relative to all those patients with at least 1 primary asthma-related visit in both time periods.
There were 4 baseline practice characteristics considered to be potential modifying variables: use of electronic health record (EHR), presence of midlevel clinician (physician assistant or nurse practitioner), clinician ownership (yes/no), and number of clinicians in the practice.
Statistical Analyses
Practice characteristics were described in terms of frequencies and proportions, with possible differences between study arms assessed using the Fisher exact test. We used the McNemar test for unadjusted comparisons of the proportions of patients receiving each recommended guideline among patients seen in both the pre- and postintervention period. We used generalized estimating equation (GEE) models to assess whether the covariate-adjusted pre- to postintervention change differed between the 3 intervention arms and the control arm for each guideline implementation measure. Modeling included study arm, study period (post- vs preintervention), the interaction of study arm with study period, and 5 covariates: presence of mid-level clinicians, clinician-owned practice, practice size, medical record type, and the practice-based research network location. The GEE approach was used to fit repeated measures models (SAS Institute, Inc), using a step-down likelihood approach to eliminate covariates that did not contribute significantly to the model. A significant interaction of study arm with study period would indicate that change over time in an outcome differed significantly among study arms.
Because the planned implementation strategies may not have been applied equally across practices, we defined adequate adherence for LLC and PF sites as follows: minimum of 3 (of 6) LLC meetings, a minimum of 12 (of a possible 24) PF visits. Sites assigned to receive both interventions had to meet minimum standards for each. To further investigate the effect of the interventions, we refit the final models developed on all practices to the data excluding those that did not adequately adhere to the assigned interventions.
The Institutional Review Boards at the University of Oklahoma Health Sciences Center and the State University of New York in Buffalo approved the study.
RESULTS
Practice Characteristics and Study Population
Of the 51 practices recruited, 45 enrolled in the study. Two practices could not participate in final medical record abstraction, leaving 43 with baseline characteristics that did not differ significantly between study arms (Table 1). Of the 2,226 patients meeting abstraction criteria, 211 had no asthma-related visits during the 27-month study period, and 12 did not include all required information, leaving 2,003 patients with 7,106 visits for whom asthma was addressed. Among these patients, 1,016 had at least 1 asthma-related visit in both the pre- and postintervention periods. Implementation of guideline recommendations was based on these 1,016 patients. The patients were demographically similar between study arms; overall 56% were female, 67% reported allergies, and 96% had no concurrent diagnosis of chronic obstructive pulmonary disease (P >.5) (data not shown).
Table 1.
Practice Characteristics by Study Arm
| Practice Characteristic | LLC | PF | PF + LLC | Control | P Valuea |
|---|---|---|---|---|---|
| Participating practices, No. | 10 | 10 | 12 | 11 | NA |
| Medical record type, No. (%) | |||||
| Electronic health record | 4 (40) | 8 (80) | 9 (75) | 5 (45) | .14 |
| Paper | 6 (60) | 2 (20) | 3 (25) | 6 (55) | |
| Mid-level practitioners, No. (%) | |||||
| Yes | 6 (60) | 6 (60) | 9 (75) | 6 (55) | .76 |
| No | 4 (40) | 4 (40) | 3 (25) | 5 (45) | |
| Practice size, No. (%) | |||||
| 1 | 3 (30) | 2 (20) | 2 (17) | 2 (18) | .67 |
| 2 | 2 (20) | 2 (20) | 3 (25) | 2 (18) | |
| 3–5 | 4 (40) | 1 (10) | 3 (25) | 5 (45) | |
| ≥6 | 1 (10) | 5 (50) | 4 (33) | 2 (18) | |
| Clinician owned, No. (%) | |||||
| Yes | 6 (60) | 5 (50) | 5 (42) | 4 (36) | .72 |
| No | 4 (40) | 5 (50) | 7 (58) | 7 (64) | |
LLC = local learning collaborative; PF = practice facilitation.
P values were obtained from the Fisher exact test of the Fisher-Freman-Halton exact test.
Matched Pairs Analyses
Implementation of all 6 guideline recommendations increased significantly postintervention in the overall study population in the unadjusted analyses (P ≤.002) (Table 2). Matched pairs analyses within each study arm are displayed in Table 3. Significant improvements were seen for 2 guideline recommendations in the control group, 3 in the PF arm, 4 in the LLC arm, and 5 in the PF + LLC arm. For descriptive purposes, a table showing the percentage change in implementation of each of the guideline recommendations by practice within intervention arm is available as a Supplemental Appendix. There was an apparent deterioration of performance of 1 guideline recommendation in 3 of the practices and deterioration of performance of 2 recommendations in 1 practice. Given the small number of asthmatic patients seen in these practices, these changes may or may not be indicative of a potential adverse effect of the interventions.
Table 2.
Unadjusted Analyses of Guideline Implementation for the Overall Study Population
| Study Population of Interest | Outcome Measure | Matched Pairs No. | Preintervention % | Postintervention % | Change % | P Valuea |
|---|---|---|---|---|---|---|
| All asthmatic patients | Assessment of asthma severity | 977 | 37 | 56 | 19 | <.001 |
| Assessment of environmental triggers | 1,016 | 38 | 52 | 14 | <.001 | |
| Asthma action plan | 983 | 7 | 11 | 4 | .002 | |
| At each asthma visit | Assessment of level of control | 937 | 40 | 56 | 16 | <.001 |
| All patients with persistent asthma | Asthma controller medications | 867 | 82 | 87 | 5 | .002 |
| Primary asthma visit | Asthma follow-up visits | 410 | 72 | 79 | 7 | .002 |
Obtained from McNemar test.
Table 3.
Unadjusted Analyses of Guideline Implementation by Study Arm of Matched Pairs
| Study Population of Interest | Outcome Measure | Control | PF | LLC | PF + LLC | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||||||
| No. | Pre % | Post % | No. | Pre % | Post % | No. | Pre % | Post % | No. | Pre % | Post % | ||
| All asthmatic patients | Assessment of asthma severity | 188 | 45 | 56a | 283 | 42 | 71a | 196 | 28 | 39a | 310 | 33 | 52a |
| Assessment of environmental triggers | 192 | 58 | 70a | 290 | 42 | 57a | 202 | 17 | 24 | 332 | 36 | 54a | |
| Asthma action plan | 180 | 13 | 15 | 282 | 7 | 10 | 199 | 3 | 8b | 322 | 7 | 11 | |
| At each asthma visit | Assessment of level of control | 172 | 44 | 45 | 276 | 55 | 73a | 276 | 24 | 45a | 301 | 34 | 54a |
| All patients with persistent asthma | Asthma controller medications | 153 | 79 | 80 | 254 | 84 | 87 | 188 | 78 | 87b | 289 | 84 | 91b |
| Primary asthma visit | Asthma follow-up visits | 66 | 62 | 70 | 111 | 78 | 83 | 78 | 65 | 72 | 155 | 74 | 84b |
LLC = local learning collaborative; PF = practice facilitation; post = postintervention; pre = preintervention.
Significantly improved by matched pairs (McNemar test), P <.01.
Significantly improved by matched pairs (McNemar test), P <.05.
Multivariate GEE Modeling
The interaction of study arm with study period contributed significantly to the assessment of severity model (P = .002) (Table 4). The pre- to postintervention improvement in assessment of severity was 2.5 times (95% CI, 1.7–3.8) higher in the PF practices compared with control practices. With respect to additional covariates, the odds of having an asthma severity assessment were approximately 40% lower for those practices with EHRs and 80% higher for practices with mid-level practitioners (odds ratio [OR] = 0.6, 95% CI, 0.5–0.8, and OR = 1.8, 95% CI, 1.3–2.4, respectively).
Table 4.
Multivariable Modeling of the Pre- to Postintervention Change in the Outcome Measures by Study Arm
| Variable | Assessment of Asthma Severity OR (95% CI) | Assessment of Environmental Triggers OR (95% CI) | Asthma Action Plan OR (95% CI) | Assessment of Level of Contro OR (95% CI) | Asthma Controller Medications OR (95% CI) | Asthma Follow-up Visits OR (95% CI) |
|---|---|---|---|---|---|---|
| Interaction of study arm with study period | ||||||
| LLC | 0.5 (0.3–0.7) | … | … | 0.8 (0.5–1.2) | … | … |
| PF | 2.5 (1.7–3.8) | … | … | 2.3 (1.5–3.5) | … | … |
| PF + LLC | 1.0 (0.7–1.5) | … | … | 1.1 (0.7–1.7) | … | … |
| Control | 1.0 | … | … | 1.0 | … | … |
| P value | .002 | .58 | . 24 | .005 | .24 | .83 |
| Medical record type | ||||||
| Electronic | 0.6 (0.5,−0.8) | 1.6 (1.1–2.2) | 0.6 (0.4–0.9) | |||
| Paper | 1.0 | 1.0 | 1.0 | |||
| P value | <.001 | .005 | .020 | … | … | … |
| Mid-level practitioners | ||||||
| Yes | 1.8 (1.3–2.4) | … | 0.5 (0.3, 0.9) | … | … | … |
| No | 1.0 | … | 1.0 | … | … | |
| P value | <.001 | … | .011 | … | … | … |
| Practice size | ||||||
| 1 | … | 0.7 (0.4–1.1) | 0.3 (0.1–0.8) | 1.0 (0.7–1.5) | … | 1.5 (0.7–3.2) |
| 2 | … | 1.9 (1.4–2.6) | 1.5 (0.9–2.4) | 1.5 (1.2–2.0) | … | 3.0 (1.6–5.5) |
| 3–5 | … | 1.6 (1.1–2.4) | 0.9 (0.5–1.7) | 0.8 (0.6–1.1) | … | 1.2 (0.7–2.1) |
| ≥6 | … | 1.0 | 1.0 | 1. 0 | … | 1.0 |
| P value | … | <.0 01 | <.0 01 | <. 0 01 | … | .001 |
| Clinician owned | ||||||
| Yes | … | 1.6 (1.3–2.1) | … | 1.5 (1.2–1.9) | … | 0.3 (0.2–0.5) |
| No | … | 1.0 | … | 1.0 | … | 1.0 |
| P value | … | <.001 | … | <.001 | … | <.001 |
| Region | ||||||
| New York | 0.6 (0.5–0.8) | 1.3 (1.0–1.6) | 0.2 (0.2–0.4) | 1.5 (1.2–1.9) | 0.7 (0.5–0.9) | 1.9 (1.3–3.0) |
| Oklahoma | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| P value | <.001 | .038 | <.001 | <.001 | .006 | .004 |
LLC = local learning collaborative; OR = odds ratio; PF = practice facilitation.
Interaction of study arm with study period also contributed significantly to assessment of level of control (P = .005), with assessment of level of control significantly higher for the PF practices compared with control practices (OR = 2.3, 95% CI, 1.5–3.5). Clinician-owned practices and those with only 2 clinicians had higher odds of improving assessment of asthma level of control (OR = 1.5, 95% CI, 1.2–1.9, and OR = 1.5, 95% CI, 1.2–2.0, respectively).
Improvement in assessment of environmental triggers did not differ by study arm (P = .58). Practices with EHRs, practices with 2 to 5 clinicians, and clinician-owned practices, however, were more likely to improve in assessment of environmental triggers (OR = 1.6, 95% CI, 1.2–2.2 for practices with EHRs; OR = 1.9, 95% CI, 1.4–2.6 for 2-clinician practices; OR = 1.6, 95% CI, 1.1–2.4 for 3- to 5-clinician practices; and OR = 1.6, 95% CI, 1.3–2.1, for clinician-owned practices).
Although change in presence of a written asthma action plan did not differ by study arm (P = .24), several covariates were associated with this outcome as well. Practices with EHRs and those with a mid-level practitioner or 1 clinician were less likely to increase their use of asthma action plans (OR = 0.6, 95% CI, 0.4–0.9; OR = 0.5, 95% CI, 0.3–0.9; and OR = 0.3, 95% CI, 0.1–0.8, respectively).
Prescription of an asthma controller medication, already high at baseline, and recommendation for an asthma follow-up visit also did not differ by study arm (P >.2). Change in controller medication prescription was not affected by practice characteristics. Practice size of 2 clinicians significantly increased asthma follow-up visit recommendations (OR = 3.0, 95% CI, 1.6–5.5), and clinician ownership decreased the odds of a follow-up recommendation (OR = 0.3, 95% CI, 0.2–0.5).
Significant differences by region were found for all outcome measures. Implementation of 3 guideline recommendations improved significantly in New York, whereas the other 3 showed significant improvement in Oklahoma. There was some variation in how the implementation strategies were implemented. For instance, practice facilitators in New York spent almost 3 times longer in each practice (3.8 hours per week compared with 1.3 hours in Oklahoma). Overall, practices in the PF + LLC study arm had fewer PF visits, on average, compared with those only receiving PF (11 vs 20), and New York practices found it more difficult to convene the full set of expected LLC sessions compared with Oklahoma practices.
To further explore implementation differences, the final models were fit to the data excluding 11 practices that did not adequately adhere to the interventions (4 LLC practices, 1 PF practice, and 6 practices assigned to PF + LLC were excluded because of inadequate exposure to PF). The models fit to the data for sites adhering to the interventions did not differ appreciably from those models fit to the full data.
The mean number of postintervention asthma-related visits differed among study arms (P <.0001). In pairwise comparisons, the mean number of postintervention visits among those in the PF arm was greater (3.8) than that of participants included in the other study arms, which ranged from 2.7 for LLC arm to 3.0 for PF + LLC arm (P <.05 using Bonferoni t tests). No differences were detected in the mean number of postintervention asthma-related visits between these other study arms (P >.05).
DISCUSSION
Participation in this quality improvement project led clinicians to better address the needs of patients with asthma. In unadjusted analyses, the intensity of the intervention appeared to correlate with effectiveness, with the control arm improving significantly on only 2 guideline recommendations, practices receiving PF improving on 3, those participating in LLC meetings improving on 4, and practices in the PF + LLC arm showing significant improvement on 5 of the 6 recommendations.
The unique aspect of this cluster randomized controlled trial was the ability to compare each intervention with the control arm, which provided additional evidence of the effectiveness of PF in motivating and supporting practice change. After controlling for covariates and comparing each intervention with a control arm, only PF proved more effective than control for 2 guideline recommendations—assessments of asthma severity and level of control. For these outcomes, PF was more effective even than PF + LLC, which was statistically no better than the control condition. This finding might be explained, in part, by less PF presence per week in PF + LLC practices, compared with PF-only practices. Although the extra visits among PF practices may have increased the probability of fulfilling the asthma guidelines for assessment of asthma severity, it does not rule out that primary care clinicians were addressing asthma more often in this study arm as a result of the efforts of the practice facilitators. The assessment of asthma level of control was not affected by the extra visits among PF practices, because this guideline was examined using all available visits.
We also examined practice-level contextual variables. Use of an EHR was associated with less improvement in assessment of asthma severity and use of asthma action plans, whereas practices with an EHR were more successful at improving assessment of environmental triggers. This finding suggests that practices are able to use EHRs to improve simple processes but have more trouble programming them to do more complex tasks like creating action plans, which argues for the need for more information technology support than provided in this project.
Presence of mid-level clinicians was expected to improve implementation of all guideline recommendations, but it was associated only with improvement in assessment of asthma severity, whereas practices with mid-level clinicians were less likely to increase their use of action plans. Region of practice was statistically associated with implementation of all 6 guideline recommendations, but the direction of effect was evenly split among recommendations, even after accounting for variation in adherence to the interventions. The interesting pattern of regional differences in the recommendations affected suggests that contextual factors—other than those included in our model—could be operating at the local or network level.
Our findings raised practical and philosophical questions about the design of interventions to improve guideline implementation. Although practices were provided with validated tools and hands-on guidance, they were free to identify target goals, determine improvement strategies, and implement changes in any sequence. More significant gains might have been achieved had practices been encouraged to follow a prescribed, logical sequence of steps. For example, practices might have had greater success if we had urged that they first establish planned visits for all asthmatic patients, then specify content of visits, create templates and order sets in the record, and finally focus on creation of action plans. The importance of sequencing of improvement strategies should be further investigated to assist in guideline dissemination and implementation.7
Limitations
There are several limitations to this study. First, our study averaged 25 patients per practice, although we estimated that 60 were required for multivariable modeling with adjustment for clustering by practice. Many patients with a diagnosis of asthma in billing records did not have evidence of asthma in the medical record or had mild disease and were only seen occasionally. Despite a much smaller sample size than anticipated, we found a statistically significant improvement among the study arms assigned a practice facilitator compared with control practices for 2 outcomes. A larger sample may have yielded additional significant findings and enabled adjustment for clustering within practices.
All outcomes measured were intermediate, process-oriented outcomes focused on practice behaviors but known to have positive impacts on patient outcomes. For example, reduced outpatient and emergency department visits, hospitalizations, and cost of asthma care have been clearly linked to higher rates of adherence to the NHLBI guidelines, particularly for use of controller medications and action plans.34,35
Inclusion of patient visits that occurred during the active intervention period may have blunted the size of the eventual measured effect of the interventions. It was also beyond the scope of our study to examine whether guideline improvements continued to improve or were sustained beyond 1 year postintervention.
It is possible that adherence to guideline recommendations could have increased simply because of the extra asthma-related visits in the PF group (more opportunities to adhere). Assessment of level of control, however, which was calculated as the percentage of visits at which this assessment was done, increased significantly in the PF group, suggesting that the interventions, rather than the number of visits, were driving the changes. We also cannot rule out the possibility that primary care clinicians were addressing asthma more often in this study arm as a result of the efforts of the practice facilitators, resulting in more asthma-related visits.
It proved challenging to ensure that LLC meetings occurred monthly with all practices in attendance, and that practices could accommodate regular PF visits. Even so, the findings remained remarkably consistent after excluding those practices not adherent to the interventions.
Finally, although the study was not designed to collect data for cost-effectiveness analysis, our experience and that of others estimate the cost of a 6-month practice facilitation intervention, including performance assessments, academic detailing, and supervision, to be between $7,500 and $15,000 per practice.21,36,37 The LLCs cost approximately one-half as much.
This study adds to the established literature showing the effectiveness of PF for at least some aspects of guideline implementation. Despite a smaller than expected sample size, we found that facilitation was significantly better than education, practical tools, and performance feedback alone in implementation of 2 of the 6 key NHLBI asthma guideline recommendations.
Footnotes
Conflict of Interest: authors report none.
Funding support: Financial support was received from the National Heart, Lung and Blood Institute, HHS-5R01HL091827.
Disclaimer: The views expressed here are those of the authors and do not necessarily reflect the views of the National Heart, Lung and Blood Institute.
Supplementary materials: Available at http://www.AnnFamMed.org/content/12/3/233/suppl/DC1/
References
- 1.Yawn BP, Yawn RA. Measuring asthma quality in primary care: can we develop better measures? Respir Med. 2006;100(1):26–33 [DOI] [PubMed] [Google Scholar]
- 2.Cabana MD, Slish KK, Nan B, Lin X, Clark NM. Asking the correct questions to assess asthma symptoms. Clin Pediatr (Phila). 2005;44(4):319–325 [DOI] [PubMed] [Google Scholar]
- 3.Diette GB, Skinner EA, Markson LE, et al. Consistency of care with national guidelines for children with asthma in managed care. J Pediatr. 2001;138(1):59–64 [DOI] [PubMed] [Google Scholar]
- 4.Diette GB, Patino CM, Merriman B, et al. Patient factors that physicians use to assign asthma treatment. Arch Intern Med. 2007;167(13):1360–1366 [DOI] [PubMed] [Google Scholar]
- 5.Apter AJ, Van Hoof TJ, Sherwin TE, Casey BA, Petrillo MK, Meehan TP. Assessing the quality of asthma care provided to Medicaid patients enrolled in managed care organizations in Connecticut. Ann Allergy Asthma Immunol. 2001;86(2):211–218 [DOI] [PubMed] [Google Scholar]
- 6.Warman KL, Silver EJ, Stein RE. Asthma symptoms, morbidity, and antiinflammatory use in inner-city children. Pediatrics. 2001;108(2): 277–282 [DOI] [PubMed] [Google Scholar]
- 7.Bender BG, Dickinson P, Rankin A, Wamboldt FS, Zittleman L, West-fall JM. The Colorado Asthma Toolkit Program: a practice coaching intervention from the High Plains Research Network. J Am Board Fam Med. 2011;24(3):240–248 [DOI] [PubMed] [Google Scholar]
- 8.McMullen A, Yoos HL, Anson E, Kitzmann H, Halterman JS, Arcoleo KS. Asthma care of children in clinical practice: do parents report receiving appropriate education? Pediatr Nurs. 2007;33(1):37–44 [PubMed] [Google Scholar]
- 9.Beauchesne MF, Levert V, El Tawil M, Labrecque M, Blais L. Action plans in asthma. Can Respir J. 2006;13:306–10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cloutier MM, Hall CB, Wakefield DB, Bailit H. Use of asthma guidelines by primary care providers to reduce hospitalizations and emergency department visits in poor, minority, urban children. J Pediatr. 2005;146(5):591–597 [DOI] [PubMed] [Google Scholar]
- 11.DePue JD, Goldstein MG, Schilling A, et al. Dissemination of the AHCPR clinical practice guideline in community health centres. Tob Control. 2002;11(4):329–335 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Katz DA, Muehlenbruch DR, Brown RL, Fiore MC, Baker TBAHRQ Smoking Cessation Guideline Study Group. Effectiveness of implementing the Agency for Healthcare Research and Quality smoking cessation clinical practice guideline: a randomized, controlled trial. J Natl Cancer Inst. 2004;96(8):594–603 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Freeborn DK, Shye D, Mullooly JP, Eraker S, Romeo J. Primary care physicians’ use of lumbar spine imaging test. J Gen Intern Med. 1997;12(10):619–625 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Bland DR, Dugan E, Cohen SJ, et al. The effects of implementation of the Agency for Health Care Policy and Research urinary incontinence guidelines in primary care practices. J Am Geriatr Soc. 2003; 51(7):979–984 [DOI] [PubMed] [Google Scholar]
- 15.Thompson C, Kinmonth AL, Stevens L, et al. Effects of a clinical-practice guideline and practice-based education on detection and outcome of depression in primary care: Hampshire Depression Project randomised controlled trial. Lancet. 2000;355(9199):185–191 [DOI] [PubMed] [Google Scholar]
- 16.Shojania KG, McDonald KMW. R. M., Owens DK. Closing the quality gap: a critical analysis of quality improvement srategies. In: AHRQ Publication No. 04(07)-0051-5; 2007 [Google Scholar]
- 17.Wilcock PM, Campion-Smith C, Head M. The Dorset Seedcorn Project: interprofessional learning and continuous quality improvement in primary care. Br J Gen Pract. 2002;52(Suppl):S39–S44 [PMC free article] [PubMed] [Google Scholar]
- 18.Bertoni AG, Bonds DE, Chen H, et al. Impact of a multifaceted intervention on cholesterol management in primary care practices. Arch Intern Med. 2009;169(7):678–686 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ornstein S, Nietert PJ, Jenkins RG, Wessell AM, Nemeth LS, Rose HL. Improving the translation of research into primary care practice: results of a national quality improvement demonstration project. Jt Comm J Qual Patient Saf. 2008;34(7):379–390 [DOI] [PubMed] [Google Scholar]
- 20.Baskerville NB, Liddy C, Hogg W. Systematic review and meta-analysis of practice facilitation within primary care settings. Ann Fam Med. 2012;10(1):63–74 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Knox L, Taylor EF, Geonnotti K, et al. Developing and running a primary care practice facilitation program: a how-to guide. AHRQ Publication No. 12-0011, December 2011 [Google Scholar]
- 22.Nagykaldi Z, Mold JW, Aspy CB. Practice facilitators: a review of the literature. Fam Med. 2005;37(8):581–588 [PubMed] [Google Scholar]
- 23.Mittman BS. Creating the evidence base for quality improvement collaboratives. Ann Intern Med. 2004;140(11):897–901 [DOI] [PubMed] [Google Scholar]
- 24.ØVretveit J, Bate P, Cleary P, et al. Quality collaboratives: lessons from research. Qual Saf Health Care. 2002; 11(4):345–351 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Cretin S, Shortell SM, Keeler EB. An evaluation of collaborative interventions to improve chronic illness care. Framework and study design. Eval Rev. 2004;28(1):28–51 [DOI] [PubMed] [Google Scholar]
- 26.Young PC, Glade GB, Stoddard GJ, Norlin C. Evaluation of a learning collaborative to improve the delivery of preventive services by pediatric practices. Pediatrics. 2006;117(5):1469–1476 [DOI] [PubMed] [Google Scholar]
- 27.Pearson ML, Wu S, Schaefer J, et al. Assessing the implementation of the chronic care model in quality improvement collaboratives. Health Serv Res. 2005;40(4):978–996 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Rogers EM. Change agents. In: Rogers, ed. Diffusion of Innovations. 5th ed New York, NY: Free Press; 2003 [Google Scholar]
- 29.Lineker SC, Bell MJ, boyle J, et al. Implementing arthritis clinical practice guidelines in primary care. Med Teacher. 2009;31:230–7 [DOI] [PubMed] [Google Scholar]
- 30.Mangione-Smith R, Schonlau M, Chan KS, et al. Measuring the effectiveness of a collaborative for quality improvement in pediatric asthma care: does implementing the chronic care model improve processes and outcomes of care? Ambul Pediatr. 2005;5(2):75–82 [DOI] [PubMed] [Google Scholar]
- 31.Aspy CB, Mold JW, Thompson DM, et al. Integrating screening and interventions for unhealthy behaviors into primary care practices. Am J Prev Med. 2008;35(5)(Suppl): S373–S380 [DOI] [PubMed] [Google Scholar]
- 32.NHLBI. National Ashma Education and Prevention Program. Guidelines for the Diagnosis and Management of Asthma. Washington, DC: NIH; 2007. Report No.: Expert Panel Report 3 [Google Scholar]
- 33.Yawn BP, Bertram S, Wollan P. Introduction of Asthma APGAR tools improve asthma management in primary care practices. J Asthma Allergy. 2008;1:1–10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Cloutier MM, Grosse SD, Wakefield DB, Nurmagambetov TA, Brown CM. The economic impact of an urban asthma management program. Am J Manag Care. 2009;15(6):345–351 [PubMed] [Google Scholar]
- 35.Bateman ED, Frith LF, Braunstein GL. Achieving guideline-based asthma control: does the patient benefit? Eur Respir J. 2002;20(3): 588–595 [DOI] [PubMed] [Google Scholar]
- 36.Hogg W, Baskerville N, Lemelin J. Cost savings associated with improving appropriate and reducing inappropriate preventive care: cost-consequences analysis. BMC Health Serv Res. 2005;5(1):20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Culler SD, Parchman ML, Lozano-Romero R, et al. Cost estimates for operating a primary care practice facilitation program. Ann Fam Med. 2013;11(3):207–211 [DOI] [PMC free article] [PubMed] [Google Scholar]