Academy Health Organizational influences on time pressure stressors - PowerPoint PPT Presentation

Kathryn McDonald, PhD, MM Academy Health Organizational influences on time pressure stressors and potential patient consequences in primary care

Acknowledgements Co-Authors: Hector Rodriguez, PhD, MPH, Stephen Shortell, PhD, MPH, MBA This project received support from: PCORI Grant IHS-1310-06821 and AHRQ Grant 1U19HS024075

Background  900 million patient visits annually in ambulatory care clinics  Professionals and patients sensitive to the potential adverse effects of having “too little time”  Patient safety concerns  Problematic for preparing patient to co-produce health  Organizations effect the amount of time that professionals have to do their work

Stressor-Stress-Performance in Ambulatory Care Context Organizational� Demands Environmental� Demands Individuals� Patient� Outcomes &� Teams Practice� Level� Efficiency� Outcomes� Time� Pressure Response� to� Time� (Stressor) Pressure Performance (Stress) Encounter� Level� Time� Pressure (Stressor) Hypothesis: P ractice level time pressure and Encounter level time pressure are two s eparate cons tructs , and as such act independently on clinic members . Adapted from: Kavanagh J. Stress and performance. A review of the literature and its applicability to the military. RAND. 2005.

Research Questions  What clinic factors are associated with Practice- Level Time Pressure?  What clinic factors are associated with Encounter-Level Time Pressure?  Are these Time Pressure constructs associated with Patient-Reported Experience of Care?

Methods: Overview  Cross-sectional, observational study  Setting:  16 randomly selected primary care practice sites in two large Accountable Care Organizations (ACOs) in Chicago and LA  Data: 2 surveys, 2 nd wave of study, surveys fielded in 2016  clinic team members (physicians, nurses, medical assistants, receptionists, diabetic educators, dieticians) [84.4% response rate]  patient with diabetes and/or cardiovascular disease (CVD) [73.5% response rate] Shortell SM, et al. A multilevel analysis of patient engagement and patient-reported outcomes in primary care practices of accountable care organizations. JGIM, 2017

Methods: Dependent Variables  Clinic analysis:  Work conditions questions included in team survey  Adapted from Linzer et al:  Practice-level Time Pressure: Practice atmosphere, chaos scale *  Encounter-level Time Pressure Effects (7 items)**  Patient level analysis  PACIC-11***, patient experience with chronic care support per Wagner et al model • * Linzer M, et al. Physician Worklife Study. JGIM, 2000 & 2015, and others • ** Personal communication with Mark Linzer • *** Glasgow RE, et al. Diabetes Care, 2005

Methods: Independent Variables  Clinic analysis (team survey)  Workgroup role  Patient-centered culture  Relational coordination among team  Health information technology capabilities and use  Leadership facilitation  Solidarity culture  Clinic site (16)  Accountable care organization (2)  Patient level analysis (patient survey)  Patient demographics  Patient Activation Measure (PAM, Hibbard et al)  CollaboRATE (Elwyn et al)

ACO� Effects Leadership� Facilitation Individual� and� Teams Workgroup� Practice- Role: Level Time� Measure:� Chaos Med� Assistant� Pressure (Perceived) vs� Other [ Stressor] Other� Clinic� Effects Response� to� Time� Performance Pressure HIT� Capability [ Stress ] Measure:� Coordination� Encounter- Patient-Reported� Capability Experience� of� Level Time� Measure:� Patient� Care� (PACIC-11) Patient� Pressure Effects� (Belief) Centeredness [ Stressor ] Solidarity� Culture Organization Responses to Environment Stressors  Team Stress Response  Patient Outcomes

Analysis  Time Pressure dependent variable metrics  Analyze missing data  Item correlations, Factor analysis, etc  Clinic level models  Multivariate logistic regression for dichotomous dependent variables  Patient level models  Hierarchical: Patients (1 st level) nested in clinics (2 nd level)

16 Primary Care Practices 90 N=274, Medical Team 80 N=353, All 70 Respondents (#) 60 50 40 30 20 10 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Clinic N, Med Team N, All

Practice-Level (Chaos) and Encounter-Level Time Pressure Not Correlated at Item Level Practice-Level (Chaos) Practice Level (Chaos) ALL MEDICAL ONLY N=228 N=205 Hypertension Dx -0.033 -0.040 Depression Dx -0.007 -0.010 Interaction Dx 0.027 0.024 Ophthalmology Dx -0.019 -0.029 Alcohol Dx -0.012 -0.012 Ace Tx -0.049 -0.051 Aspirin Tx -0.034 -0.039

Two Time Pressure Stressors: Variation By Clinic 100.00 Respondents Perceiving Stressor (%) 80.00 60.00 40.00 20.00 0.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Clinic Encounter Level, Med Encounter Level, All Practice Level (Chaos), Med Team Practice Level (Chaos), All

ACO Effects Encounter-Level Practice-Level Time Pressure Time Pressure (Chaos) Odds Ratio P >| z | Odds Ratio P >| z | ACO A Reference Reference ACO B 1.087 0.757 0.523 0.027

Odds Ratio P>|z| Leadership 0.922 0.000 Clinic 1 Reference 2 0.152 0.009 3 0.181 0.095 4 0.195 0.251 Practice- 5 0.048 0.011 6 0.175 0.033 Level Time 7 1.916 0.609 8 0.097 0.012 9 0.048 0.024 Pressure: 10 0.052 0.009 11 0.151 0.012 Hypotheses 12 0.708 0.719 13 0.135 0.005 Tests 14 0.349 0.283 15 0.105 0.022 16 0.095 0.006 Physician Reference Diabetes Educator 1.162 0.844 Medical Assistant 2.269 0.033 Nursing 1.589 0.292

Encounter-Level Time Pressure: Hypotheses Tests Odds Ratio P>|z| Patient Centered 0.829 0.000 Health IT 0.433 0.002 Relational Coordination 0.984 0.000 Solidarity 1.059 0.277

Including All Clinic Variables: Testing for Anticipated Null Effects  Practice-Level (Chaos) Time Pressure  No unanticipated effects  Encounter-Level Time Pressure  Unanticipated effects --  3 clinics significantly different  Lower odds of perceiving time pressure effects on care by medical assistants and nurses, compared to physicians

Patient Effects: PACIC-11 (N=1169) Practice-Level Time Pressure (Chaos) 0.743 ** (0.072) Encounter-Level Time Pressure 0.828 (0.136) Good English Proficiency (vs poor) 0.930 0.919 (0.057) (0.057) 18-24 (reference) 1.390 1.419 25-44 (0.433) (0.442) 1.339 (0.405) 1.347 45-64 (0.408) 1.309 1.315 65+ (0.396) (0.398) Female 0.900 *** (0.028) 0.903 ** (0.028) 8 grade or less (reference) GED or some high school 0.831 ** (0.057) 0.827 ** (0.057) 4 yr college degree or some college 0.734 *** (0.053) 0.727 *** (0.053) More than 4 year college degree (0.061) 0.749 *** 0.740 *** (0.060) Social Functioning (1-5) 0.986 0.988 (0.022) (0.022) Physical Functioning (1-5) 0.985 0.983 (0.025) (0.025) Emotional Functioning (1-4) 1.040 1.039 (0.028) (0.028) Patient Activation Measure (PAM) 1.281 *** (0.051) 1.280 *** (0.051) Other Providers Involved (vs Dr Only) 1.102 ** (0.036) 1.101 ** (0.036) CollaboRATE (mean) 1.658 *** (0.029) 1.655 *** (0.029) Exponentiated coefficients (Standard errors) * p < 0.05, ** p < 0.01, *** p < 0.001

Limitations  Cross-sectional, so causality unknown  Encounter-level time pressure metric only provides perceptions about likelihood of missing important diagnostic and treatment opportunities since it has not been verified by medical record review  Risk of spurious associations due to common methods bias for clinic analysis, but not patient analysis

Conclusions/ Implications  Two measures used in this study seem to be capturing distinct processes and experiences    Practice-level time pressure     Encounter-level time pressure  Findings suggest organizational leverage for managing experiences and perceptions of time pressure for health care systems, such as ACOs    Leadership facilitation important for practice-level time pressure     HIT capability, patient-centered culture and relational coordination potentially important for encounter-level time pressure  Potential time pressure effects on patients  Patient safety – missing diagnostic and treatment opportunities  Patient support – not equipping patient adequately to care for themselves outside of clinic visits