Article Figures & Data
Figures
- Figure 1.
Practice Survey Question About Disruptions.
- Appendix Figure 1.
Variation in facilitation hours by practice
Tables
Characteristics Practice Facilitation (Mean hours [SD]) 18.2 (16.7) Practice Size/Ownership, n (%) Group clinician-owned 248 (25.1) Solo clinician-owned 174 (17.6) Hospital/HS/HMO 266 (27.0) Safety net 243 (24.6) Other 23 (2.3) Missing 33 (3.3) Practice Location, n (%) Large town 152 (15.4) Rural area 175 (17.7) Suburban 72 (7.3) Urban 588 (59.6) Participation in Other Demonstration Projects, n (%) No 619 (62.7) Yes 316 (32.0) Missing 52 (5.3) Cooperative, n (%) Midwest 136 (13.8) North Carolina 134 (13.6) Northwest 175 (17.7) New York City 90 (9.1) Oklahoma 196 (19.9) Southwest 181 (18.3) Virginia 75 (7.6) Abbreviations: HS, Health System; HMO, Health Maintenance Organization; SD, standard deviation.
- Table 2.
Disruptive Events Reported by Practices Before and During EvidenceNOW Interventions
Before Intervention During Intervention N (%) N (%) Number of Disruption Types 0 485 (49.1) 476 (48.2) 1 329 (33.3) 318 (32.2) 2 125 (12.7) 146 (14.8) 3+ 48 (4.9) 47 (4.8) Disruption Type Lost clinician 279 (28.3) 318 (32.2) Lost office manager/head nurse 164 (16.6) 212 (21.5) Organizational* 52 (5.3) 59 (6.0) Information system† 142 (14.4) 75 (7.6) - Table 3.
Association of Disruptive Events During Intervention with Participation in Facilitation
Facilitation Participation (hours, 95% CI)* Difference (hours, 95% CI) P value One or More Disruptions Present Yes No 14.9 (8.3, 21.6) 14.5 (9.0, 20.1) 0.39 (−1.18, 1.96) 0.630 Disruption Type Present Yes No Lost clinician 16.4 (8.5, 24.3) 15.1 (8.9, 21.6) 1.28 (−1.00, 3.56) 0.388 Lost office manager/head nurse 15.9 (8.7, 23.2) 15.3 (8.5, 22.2) 0.60 (−1.51, 2.70) 0.786 Organizational 14.5 (7.1, 21.9) 15.5 (8.7, 22.3) −1.01 (−2.58, 0.57) 0.293 Information system 14.4 (7.1, 21.8) 15.6 (8.6, 22.6) −1.17 (−3.26, 0.93) 0.393 Abbreviation: CI, confidence interval.
↵* Participation in facilitation shown as estimated mean hours, adjusted for events at baseline, practice size/ownership, practice location, participation in other demonstration projects, and Adaptive Reserve. Type III test utilized, testing if all 6 comparisons are equivalent to zero. Statistical significance defined as p-value < 0.05.
- Appendix Table 1.
Descriptions of “Other” Responses to Post-Intervention Practice Survey Question About Disruptions
“Other” Write-in Responses as Categorized Baseline Survey (n) Post-Intervention Survey (n) Methodological Treatment Implemented a new or different EHR 2 0 Recoded to existing question response Lost one or more clinicians 2 3 Recoded to existing question response Lost one or more office managers or head nurses 0 1 Recoded to existing question response Moved to a new location 0 2 Recoded to existing question response New billing system 2 2 Recoded to existing question response Ownership change 1 1 Recoded to existing question response New staff 24 41 Excluded because of varied and ambiguous nature of “staff” and lack of clarity if new staff were additions or replacements to existing staff Lost staff 2 23 Excluded because of varied and ambiguous nature of “staff” New EHR features and EHR challenges 5 6 Excluded because of small n Practice expansion/merger not involving ownership change 11 10 Excluded because of small n; unclear if these are the same as existing question response Practice/site closed 0 1 Excluded because of small n Miscellaneous 15 17 Excluded because of small n and heterogeneity of responses Total responses excluded from analysis 64 106 Abbreviation: EHR, electronic health record.
- Appendix Table 2.
Estimated Marginal Hours and Difference in Hours (95% CI) of Participation in Practice Facilitation by Occurrence of Disruption*
Difference P value Disruption Types No Yes Lost clinician 15.1 (8.9, 21.6) 16.4 (8.5, 24.3) 1.28 (−1.00, 3.56) 0.388 Lost office manager 15.3 (8.5, 22.2) 15.9 (8.7, 23.2) 0.60 (−1.51, 2.70) 0.786 Purchased/new affiliation/new location 15.5 (8.7, 22.3) 14.5 (7.1, 21.9) −1.01 (−2.58, 0.57) 0.293 New EHR or billing software 15.6 (8.6, 22.6) 14.4 (7.1, 21.8) −1.17 (−3.26, 0.93) 0.393 0 events 1+ events One or More Disruptions 14.5 (9.0, 20.1) 14.9 (8.3, 21.6) 0.39 (−1.18, 1.96) 0.630 Number of Disruptions† 0 events 1 event 15.2 (8.9, 21.5) 15.5 (7.7, 23.3) 0.32 (−1.93, 2.58) 0.983 0 events 2 events 15.2 (8.9, 21.5) 15.7 (8.8, 22.7) 0.57 (−2.35, 3.50) 0.958 0 events 3+ events 15.2 (8.9, 21.5) 15.2 (8.3, 22.2) 0.07 (−1.93, 2.07) 1.000 1 event 2 events 15.5 (7.7, 23.3) 15.7 (8.8, 22.7) 0.25 (−2.72, 3.22) 0.996 1 event 3+ events 15.5 (7.7, 23.3) 15.2 (8.3, 22.2) −0.26 (−2.00, 1.48) 0.981 2 events 3+ events 15.7 (8.8, 22.7) 15.2 (8.3, 22.2) −0.51 (−2.82, 1.81) 0.943 Abbreviation: CI, confidence interval; EHR, electronic health record.
↵* Estimated marginal hours are adjusted over the levels of the following variables: baseline disruptions, practice size ownership, practice location, other demonstration projects, and Adaptive Reserve.
↵† Difference not included for all 6 comparisons. Also, Type III test utilized, testing if all 6 comparisons are equivalent to zero.
- Appendix Table 3.
Association of Disruptions During Intervention with Participation in Practice Facilitation: Complete Model, with Imputed and Non-Imputed Results.*
Estimate (95% CI) Imputed Estimates (95% CI) Intercept 17.76 (13.04, 24.20) 17.52 (12.50, 24.56) Events During Intervention 0 events Reference Reference 1 + event 1.03 (0.93, 1.13) 1.03 (0.92, 1.15) Events Before Intervention 0 events Reference Reference 1 + event before intervention 1.12 (1.03, 1.22) 1.12 (1.03, 1.22) Practice Size/Ownership Group clinician-owned Reference Reference Solo clinician-owned 0.92 (0.76, 1.12) 0.92 (0.76, 1.12) Hospital/HS/HMO 0.83 (0.71, 0.97) 0.84 (0.71, 0.98) Safety net 0.77 (0.54, 1.09) 0.77 (0.53, 1.12) Other 0.74 (0.63, 0.87) 0.75 (0.62, 0.90) Missing 0.90 (0.68, 1.19) - Practice Location Large town Reference Reference Rural area 0.96 (0.84, 1.11) 0.96 (0.83, 1.12) Suburban 0.73 (0.54, 0.97) 0.73 (0.53, 0.99) Urban 1.10 (0.79, 1.54) 1.10 (0.77, 1.58) Participate in Other Initiatives No Reference Reference Yes 0.99 (0.87, 1.11) 0.99 (0.88, 1.13) Missing 0.99 (0.80, 1.23) - Adaptive Reserve Low Reference Reference High 0.99 (0.89, 1.10) 1.00 (0.89, 1.12) Missing 0.87 (0.78, 0.97) - Abbreviations: CI, confidence interval; HMO, Health Maintenance Organization; HS, Health System.
↵* Reference groups (the intercept) include zero follow-up disruptions, zero baseline disruptions, group clinician-owned, large town, no other demonstration projects, and low Adaptive Reserve.
- Appendix Table 4.
Alternative Definitions of Participation in Practice Facilitation: Summary and Association with Any Disruption
Name Definition Summary Association With Practice Event* (95% CI) Consistency of Facilitation Visits, Mean (SD) Months during intervention period with an in-person PF visit 7.61 (3.22) 0.86 (0.59, 1.31) Length of Facilitation Visit, Mean (SD) Total hours of facilitation/total number of in-person PF visits 1.63 (0.89) 1.00 (0.94, 1.08) Dose Categories, N (%) 0.94 (0.74, 1.18) Low: <10 hours 384 (38.9) Short: ≥10 -<50 hours, <10 months 188 (19.0) Consistent: ≥10-<50 hours, ≥10 months 351 (35.6) High: ≥50 hours 64 (6.5) Abbreviations: CI, confidence interval; PF, practice facilitation; SD, standard deviation.
↵* For the participation in practice facilitation outcomes, consistency of practice facilitation and length of practice facilitation visit, a Poisson regression and Gamma regression models were implemented, exploring the effect of any disruption against no disruptions. For dose categories, we used a multinomial logistic regression. Incident rate/odds ratios (exponentiated coefficients) and 95% CIs were reported.
For months with an in-person encounter, we see a 14% decrease in practice facilitation if a practice had a disruption versus no disruptions. For length of practice facilitation, we see no effect if a practice had a disruption versus no disruptions. For the pooled dose, there is a 6% decrease in the odds of a disruption occurring for a Low, Short, or Consistent dose versus a High dose.
