Hospitalization for acute medical illness is associated

Transcription

1 Original Research Impact of an Educational Intervention in a Post-Discharge VTE Prophylaxis Quality Improvement Initiative Sara L. Merwin, MPH, Haisam Ismail, MD, Nina Kohn, MBA, MA, Saima I. Chaudhry, MD, MSHS, and David J. Rosenberg, MD, MPH ABSTRACT Objective: To assess provider attitudes toward receiving an educational intervention using a human alert encouraging post-discharge VTE prophylaxis, to assess how new information was integrated into provider practice, and to determine if there were persistent effects. Methods: Medical inpatients awaiting discharge were screened to identify those at high risk of VTE who were not ordered to receive thromboprophylaxis after discharge. Eligible patients were randomized to alert or no alert. We contacted providers of patients in the alert group with the suggestion to continue thromboprophylaxis using a modified academic detailing approach, and we characterized provider attitudes towards receiving the alert. After the study was completed, we surveyed providers to assess potential effects of the alert intervention. Results: 120 provider alerts to 56 providers were studied. For the initial contact, provider responses to the alert were positive (64%), neutral (25%), dismissive (4%), and no response (7%). Across all encounters, differences were noted between specialties and provider type (p = 0.002). Survey responses for 42/56 (75%) providers were obtained, with 57% reporting learning something new and 27% reporting they had changed practice as a result of the alert. 82% said they would consider post-discharge VTE in the future. Conclusion: Our findings support that one size does not fit all when offering education to providers but suggest that academic detailing can be used successfully to stimulate critical thinking about new medical information. Provider characteristics, culture, and external factors should be considered when tailoring education efforts. Hospitalization for acute medical illness is associated with an increased risk of VTE that persists in the months following discharge [1,2]. Evidence-based guidelines recommend the use of VTE prophylaxis in hospitalized patients, but the optimal duration of prophylaxis remains unknown. Currently, the use of post-discharge thromboprophylaxis is inconsistent [3]. Although extended and post-hospital thromboprophylaxis is recommended in certain surgical patient populations, the American College of Chest Physicians (ACCP) does not specifically address extended or post-discharge prophylaxis in medically ill patients in their guidelines, with the exception of patients with cancer who are both medically ill and immobilized [4]. Studies evaluating extended VTE prophylaxis in medical patients following hospital discharge show some benefit but with an increased risk of bleeding [5 7]. Future studies, some already underway (APEX) [8], will likely focus on the highest-risk patients, in whom the benefits of VTE prophylaxis may outweigh bleeding risk. Studies have shown that physicians differ in their willingness to accept interventions and that there are many barriers to bringing new information into practice. There is evidence that targeted one-on-one education tailored to physicians can be an effective method of changing physician behavior [9,10]. A complex interplay of intrinsic and extrinsic factors, including learning styles, culture, specialty, training level, age, and practice context, influences acquisition of new knowledge. CME guidelines developed by the ACCP summarize the science of learning and knowledge transfer among physicians. Key tenets for success included repeated, sustained educational efforts and active learning strategies, including one-on-one interactions such as academic detailing [11]. Although From the North Shore-LIJ Health System, Manhasset, NY. Vol. 20, No. 10 October 2013 JCOM 447

2 VTE prophylaxis Table 1. Inclusion, Exclusion Criteria and Risk Assessment Instrument Inclusion Criteria: Patient age 18 years Planned discharge within 48 hours Cumulative VTE risk score 4 On a medical service Exclusion Criteria: Full anticoagulation therapy planned upon discharge Minor (Low) Risk Factors (1 point each): Advanced age (> 70 years) Bedrest/immobility Obesity (BMI > 29 or the presence of the word obesity in admission exam notes) Female hormone replacement therapy or oral contraceptives Major (High) Risk Factors (3 points each): Cancer (active) Prior VTE Hypercoagulability Adapted from reference 15. detailing had its origins in the pharmaceutical industry where it was employed by sales representatives, it has been adopted and adapted by the medical profession as a strategy to educate providers and improve processes of care [9,12 14]. Some of the key techniques of academic detailing are (1) conducting interviews to investigate baseline knowledge and motivations for current prescribing patterns, (2) focusing programs on specific categories of physicians as well as on their opinion leaders, (3) defining clear educational and behavioral objectives, (4) establishing credibility through a respected organizational identity, referencing authoritative and unbiased sources of information, and presenting both sides of controversial issues, (5) stimulating active physician participation in educational interactions, (6) using concise graphic educational materials, (7) highlighting and repeating the essential messages, and (8) providing positive reinforcement of improved practices in follow-up visits [10]. Our institution participated in a multicenter randomized control trial (RCT) to test whether a hospital staff member s thromboprophylaxis alert to an attending physician before discharge would increase the rate of extended out-of-hospital prophylaxis. During and after our participation, we conducted our own investigation to gain insight into how a modified academic detailing approach was received and if it was integrated into practice or had persistent or sustained effects. We hypothesized that physicians and other providers would resist an intervention to discharge patients on prolonged VTE prophylaxis since it was not a standard of care. We present here an investigation in 2 phases: (1) the impact of the initial alert on provider response and attitudes, and (2) the results of a survey assessing provider knowledge acquisition and retention. METHODS The RCT The multicenter quality improvement RCT by Piazza et al [3] was conducted at 18 study sites from April 2009 through January Medical inpatients awaiting discharge within 48 hours were identified and subsequently screened for eligibility using a previously developed assessment tool with a weighted point system (Table 1) [15,16]. Patients with point scores of 4 or greater were deemed to be at continued elevated risk for developing VTE postdischarge. This subset of patients was randomized by computer by a 1:1 ratio to the alert group or the control group. Providers of patients in the alert group were contacted during daytime work shifts, but without regard to rounding schedules, teaching responsibilities or office hours. Alerts were performed by trained study personnel (attendings, medical residents and a research coordinator). Alerts were executed in person, by page or by telephone. A scripted message was read to encourage postdischarge VTE prophylaxis for the patient in question, with an offer to send peer-reviewed journal articles with evidence to support this practice. A 48-hour follow-up to review charts was performed to determine if VTE prophylaxis discharge orders were written. Additionally as part of the multicenter trial, 90-day followup to review patient outcomes was undertaken to assess the study endpoints of mortality, thromboembolic events, and bleeding. Providers for both alert and no alert groups were queried by mail, , and telephone to obtain this information, further serving as a reminder of the alert process for the subset of providers previously contacted in the intervention group. The results of the multicenter trial, which investigated whether the alert methodology resulted in a decrease in thromboembolic events, have been summarized and published [3]. Our Study During delivery of the alert, we assessed provider reaction to receiving the information about continued VTE 448 JCOM October 2013 Vol. 20, No. 10

3 Original Research Provider Reaction Follow-up Survey to the Discharge Alert Quality Initiative Hello. This a brief follow-up survey regarding VTE prophylaxis post-discharge. You have been previously contacted regarding this matter for one (or more) of your patients in the past few months. This survey is to gauge physician reaction to the study and to see if current physician practice has been affected. 1. Since the initial contact or telephone contact during the Discharge Alert study, have you changed your practice with regard to post-discharge prophylaxis? Yes No Comment 2. If you decided against post-discharge VTE prophylaxis for your patient(s) when contacted during the study, have you thought more about the idea or would you consider adding VTE prophylaxis such as Lovenox, heparin, or PAS to any one of your patients if appropriate in the future? Yes No, because: a. There is no evidence or guidelines for this practice b. Never heard of any of my colleagues do it--there is a lack of education about this c. It was not appropriate for my patient (ie, ow platelets) d. Iinsurance (Lovenox is expensive) or it was impracticable such as PAS stockings e. It is assumed that after discharge patients are back at their baseline f. Patient compliance--they will not want to stick themselves especially with hep SQ 3x/day 3. After being notified about our study, did you read any literature about VTE prophylaxis post-discharge and are you aware of the studies regarding this such as the EXCLAIM study? Yes If so, which No 4. Did you learn anything new regarding post-discharge VTE prophylaxis? Yes (did not know about this until now) No (knew about this but evidence is weak) Figure 1. Survey tool to assess multiple effects of new education via alert methodology in a quality improvement initiative. prophylaxis. We categorized 4 possible attitude responses to the alert content qualitatively: positive (appreciation for the suggestion or a commitment to write orders), neutral, negative/dismissive, and did not respond to call or page. We also asked several short questions. We asked if the provider had prior knowledge or seen evidence supporting post-discharge thromboprophylaxis. Responses were classified as aware, not aware, and other. If aware, we asked how frequently the provider has sent medically ill persons home on prophylaxis. A request by the provider for information or an acceptance of an offer of evidence-based information was considered a proxy measure of engagement and an interest in learning new information. Provider characteristics (specialty, gender, and training level) were collected after the encounter to describe encounter responses. Several months after the conclusion of the study, we followed up with our providers who had received an alert during the study. A survey tool was created to assess multiple effects of the alert. The object of the survey was to gain insight into how information delivered in the alert was integrated into practice; if the intervention stimulated critical thinking about the need for post-discharge prophylaxis among highrisk medically hospitalized patients; and if effects were sustained. The survey instrument also captured obstacles and objections. The survey tool is shown in Figure 1. Local IRB approval was obtained to participate in the RCT and to study provider attitudes towards the alert and to follow up with providers who had been alerted during the quality initiative to ascertain knowledge acquisition and retention. Data Analysis For the first phase of the study (the alert encounter), data were analyzed in 2 ways: with the encounter as Vol. 20, No. 10 October 2013 JCOM 449

4 VTE prophylaxis the unit of analysis, and the provider as the unit of analysis. In the case of providers contacted more than once, we examined only the first encounter. Summary statistics are given as mean and standard deviation for continuous variables and frequency and percent for categorical variables. Associations between categorical variables and continuous variables were examined using the Mann-Whitney test. Associations between pairs of categorical variables were examined using the chisquare test or Fisher s exact test, as appropriate. SAS and SPSS software were used to analyze the data for this study. RESULTS A total of 510 medical inpatients were screened for eligibility. Of these, there were a total of 285 medical inpatients who met inclusion criteria. 146 patients were randomized to no alert (control), and 139 to the alert group. Of the 139 randomized alert encounters, 19 alerts occurred prior to IRB approval to study attitudes and previous knowledge about discharge VTE prophylaxis, leaving a total of 120 alerts encounters available for study (Figure 2). In the alert group of 120 patients, 99 (82.5%) patients did not receive an order for any type of thromboprophylaxis, 14 (11.7%) were prescribed either mechanical or chemoprophylaxis at discharge, 4 (3.3%) patients expired on day of discharge, and 3 (2.5%) were identified with existing VTE after the alert and were being treated. In the no alert group, 7.9% of patients were prescribed thromboprophylaxis. There was no statistically significant difference between the alert and no alert groups for orders written. Provider Attitudes Table 2 presents the qualitative attitude response to the initial alert by specialty/training and provides data about provider knowledge about evidence for VTE prophylaxis at hospital discharge and response to the offer of literature to substantiate rationale for discharge prophylaxis. For the initial contact with each provider, 36 (64%) were positive, 14 (25%) were neutral, 2 (4%) were dismissive and 4 (7%) did not respond. On initial alert encounter, when responses were dichotomized: males were more likely than females to be accepting/appreciative (p = 0.01). Since it was possible for providers to receive alerts for multiple patients over the course of the study, we examined provider responses for all alerts: There were 120 provider alert encounters or attempts, 60% with oncologists, 30% with internists and the remaining were with cardiologists, oncology fellows, and nurse practitioners. Table 3 summarizes the qualitative attitude responses for all of the alert interventions. Among all responses in all specialties and training levels, 70 (58%) were positive, 29 (24%) were neutral, 2 (2%) were dismissive and 19 (16%) did not respond. There was a statistically significant difference between specialties for the qualitative response (p = 0.002). During the alerts, the most common objections to post-discharge pharmacologic prophylaxis cited were concern about thrombocytopenia and active bleeding. We noted that some providers exhibited different attitudes across multiple responses, with no apparent pattern (Figure 3). Survey Results A total of 56 providers were surveyed several months after the completion of the study via phone calls and pages. Survey responses for 42 of these providers were obtained. The remaining providers did not return phone calls, pages or s. 42 out of 56 (75%) attendings, fellows and nurse practitioners were contacted after the completion of the alert and follow-up phases of the study. Of the 42 respondents, 25 providers had been alerted once, 4 had been alerted twice, 7 had been alerted 3 times, with the remaining 6 providers being alerted between 4 and 10 times for successive patients. Internists were more likely to respond to the survey than oncologists (p = 0.015) No other differences between respondents and nonrespondents were observed (Table 4). 57.1% of providers contacted reported they learned something new or were reminded about post discharge prophylaxis. When asked if they had changed practice as a result of the initial contact or series of contacts, 11 (26.8%) reported that they had made changes. However, several respondents did not recall being contacted regarding discharge VTE prophylaxis. When asked if they had thought more about the idea or would consider adding mechanical or pharmacologic prophylaxis in the future, 31 out of 38 (81.6%) responded affirmatively. Table 5 summarizes post-study survey results. DISCUSSION VTE is a well-known cause of medical morbidity and mortality in hospital patients. Optimizing prophylaxis against VTE in the hospital setting has therefore become an indispensable part of medical care. However, the optimal duration of therapy is still currently unknown. 450 JCOM October 2013 Vol. 20, No. 10

5 Original Research 510 high-risk medically hospitalized adult patients screened for eligibility for inclusion in discharge alert multisite quality initiative, Obtain post-discharge follow up for 285 patients to determine if provider ordered thromboprophylaxis after discharge 285 patients met inclusion criteria on risk assessment tool 139 patients randomized to alert (intervention) 146 patients randomized to "no alert" (control) 139 alert encounters with 56 providers of patients on medical service to encourage post-discharge thromboprophylaxis 19 alert encounters occurred prior to start of nested study of attitudes towards alert delivering new information Obtain 90-day followup for 285 patients to determine endpoints: DVT, PE, bleeding, mortality 120 alert encounters studied for attitudes towards receiving new information and existing knowledge about postdischarge thromboprophylaxis Follow-up survey attempted to 56 providers to assess knowledge retention and behavior change Collect demographic data on 56 providers 42 respondents to post-study survey Figure 2. Work flow diagram for quality improvement initiative procedures and 2 educational assessment substudies. There are many factors affecting provider knowledge acquisition and retention of new information. Academic detailing has been shown to be a successful strategy to improve outcomes in medical training. Specifically, engagement, rapport building and interaction may enhance knowledge transfer. We utilized a survey tool to assess whether a modified AD intervention during the course of a VTE quality improvement initiative promoted adoption of post discharge VTE prophylaxis and affected provider attitudes towards new and emerging therapies. There are no current guidelines on post-discharge VTE prophylaxis; therefore educating physicians and other providers on a practice that is being studied and investigated in current trials was challenging. A little more than Vol. 20, No. 10 October 2013 JCOM 451

6 VTE prophylaxis Table 2. Provider Attitude Towards Alert, Awareness, and Engagement (Initial Encounters for Each Provider; n = 56) Attitude Towards New Information During Alert, n (%) Aware of Lit on D/C Proph, n (%) Lit Request or Accept Offer, n (%) Positive Neutral Dismissive Did Not Respond Internal Medicine 15 (71.4) 5 (23.8) 1 (4.8) 0 (0.0) 11 (55.0) 3 (14.3) Oncology 13 (54.2) 6 (25.0) 1 (4.2) 4 (16.7) 13 (68.4) 6 (30.0) Cardiology 3 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 2 (66.7) 2 (66.7) Fellow 4 (66.7) 2 (33.3) 0 (0.0) 0 (0.0) 3 (60.0) 1 (16.7) Nurse Practitioner 1 (50.0) 1 (50.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Total 36 (64.3) 14 (25.0) 2 (3.6) 4 (7.1) 29 (59.2) 12 (23.1) Note: Sample sizes vary due to non-response. Table 3. Provider Attitude Towards Alert, Awareness, and Engagement for All Encounters (n = 120) Attitude Towards New Information During Alert, n (%) Aware of Lit on D/C Proph, n (%) Lit Request or Accept Offer, n (%) Positive Neutral Dismissive Did Not Respond Internal Medicine 29 (80.6) 6 (16.7) 1 (2.8) 0 (0.0) 21 (65.6) 5 (14.3) Oncology 33 (45.8) 19 (26.4) 1 (1.4) 19 (26.4) 42 (84.0) 8 (15.1) Cardiology 3 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 2 (66.7) 2 (66.7) Fellow 4 (66.7) 2 (33.3) 0 (0.0) 0 (0.0) 3 (60.0) 1 (16.7) Nurse Practitioner 1 (33.3) 2 (66.7) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Total 70 (58.3) 29 (24.2) 2 (1.7) 19 (15.8) 68 (73.9) 16 (16.0) Note: Multiple encounters with providers possible for different patients. Sample sizes vary due to non-response. The attitude response was statistically significant (p = 0.002) for the different specialties. half of physicians contacted reported that they learned something new. One-third of providers contacted reported that they read literature about post discharge VTE prophylaxis. Approximately one-quarter reported having changed their practice. This may be a consequence of the educational objective still being reviewed in trials and not currently part of medical practice. However, critical thinking was promoted as evidenced by the 82% of providers reporting that they would consider changing practice in the future. In the same manner that a crossover effect may have contaminated the differences between the intervention and control groups (which is not the subject of this investigation), the investigators believe that particularly among providers contacted multiple times for different patients, although the index patient(s) may not have received orders for thromboprophylaxis at discharge, subsequent patients with appropriate indications, may have benefited from the intervention to these providers. Disparate reactions noted in this study suggest varied learning styles and response patterns among internists versus specialist physicians. Audience characteristics, culture and external factors must be considered when tailoring educational efforts to providers. Understanding the provider response to new information is essential to integrating new findings into practice. The variable attitudinal responses of the same providers at different time points underscores the need to evaluate which strategies will be most successful when offering new information. We hypothesized that based upon the paucity of highlevel evidence to support post-discharge prophylaxis there would be resistance to the alert and to incorporating new knowledge in practice. Our results were equivocal; although not all providers embraced the offered suggestion, overall reception and attitude to hearing new information was well-received. The study design may have contributed to confusion about the alert message, since some patients were not suitable candidates for pharma- 452 JCOM October 2013 Vol. 20, No. 10

7 Original Research Figure 3. Varied attitude responses over time by selected providers to a modified AD alert intervention for extended VTE prophylaxis for different patients. This figure is restricted to providers who received more than a single alert. Each row represents the attitude response of a single provider. cologic prophylaxis and the evidence for mechanical prophylaxis has not been well substantiated. However on balance, during the alert encounter, most responses were favorable. Further, and notably, the post study survey results indicate that critical thinking was stimulated and that a significant proportion of providers would consider changing practice based upon having participated in dialog around the rationale for post-discharge VTE prophylaxis. CONCLUSION Our findings suggest that a modified academic detailing educational intervention can positively influence receptivity towards new and emerging VTE prophylaxis strategies. Our findings support that one size does not fit all when offering education to medical providers. Further studies are needed to elucidate what factors contribute to the integration of a novel educational message. Future investigations should examine learning differences among Table 4. Post-Study Survey Results: Specialty and Licensure of Respondents and Nonrespondents (n = 56) Respondents n = 42 Nonrespondents n = 14 Mean years since 13.2 ± ± 12.6 licensure ± SD Male 26 (78.8) 7 (21.2) Female 16 (69.6) 7 (30.4) Internal Medicine 19 (90.5) 2 (9.5) Oncology 14 (58.3) 10 (41.7) Cardiology 3 (100.0) 0 (0.0) Fellow 4 (66.7) 2 (33.3) Nurse Practitioner 2 (100.0) 0 (0.0) Note: Sample sizes vary due to non-response. Response rates were significant for internal medicine vs. oncologists, p = Vol. 20, No. 10 October 2013 JCOM 453

8 VTE prophylaxis Table 5. Post-Study Survey Results: Reported Effects From Alert Encounters by Provider Specialty Changed Practice After Intervention, n (%) Would Change Practice in Future, n (%) Learned New Information as a Result of Intervention, n (%) Read Literature Since Intervention, n (%) Internal Medicine 8 (42.11) 16 (88.9) 11 (57.9) 4 (21.1) Oncology 2 (14.3) 11 (91.7) 8 (57.1) 6 (42.9) Cardiology 0 (0.0) 1 (33.3) 2 (66.7) 1 (33.3) Fellow 1 (33.3) 2 (50.0) 2 (50.0) 2 (50.0) Nurse Practitioner 0 (0.0) 1 (100.0) 1 (50.0) 1 (50.0) Total 11 (26.8) 31 (81.6) 24 (57.1) 14 (33.3) Note: Percentages are within provider specialty. Sample sizes vary due to non-response. specialties and training level, which methods are most successful, and other facets of knowledge transfer. Corresponding author: Sara L. Merwin, MPH, LIJ Medical Center, th Ave., New Hyde Park, NY Financial disclosures: None. Author contributions: conception and design, SLM, HI, SIC, DJR; analysis and interpretation of data, SLM, NK, SIC; drafting of article, SLM, HI, SIC, DJR; critical revision of the article, SLM, NK, SIC; statistical expertise, SLM, NK; obtaining of funding, DJR; administrative or technical support, DJR; collection and assembly of data, SLM, HI. References 1. Hull RD, Schellong SM, Tapson VF, et al. Extended-duration thromboprophylaxis in acutely ill medical patients with recent reduced mobility: methodology for the EXCLAIM study. J Thromb Thrombolysis 2006;22: Spencer FA, Lessard D, Emery C, et al. Venous thromboembolism in the outpatient setting. Arch Intern Med 2007;167: Piazza G, Anderson FA, Ortel TA, et al. Randomized trial of physician alerts for thromboprophylaxis after discharge. Am J Med 2013;126: Geerts WH, Bergqvist D, Pineo GF, et al. Prevention of venous thromboembolism: American College of Chest Physicians evidence-based clinical practice guidelines. 8th ed. Chest 2008;133(6 Suppl):381S-453S. 5. Hull RD, Schellong SM, Tapson VF, et al. Extendedduration venous thromboembolism prophylaxis in acutely ill medical patients with recent reduced mobility: a randomized trial. Ann Intern Med 2010;153: Roser-Jones C, Becker RC. Apixaban: an emerging oral factor Xa inhibitor. J Thromb Thrombolysis 2010;29: Cohen, AT, Spiro TE, Büller HR, et al. Extended-duration rivaroxaban thromboprophylaxis in acutely ill medical patients: MAGELLAN study protocol. J Thromb Thrombolysis 2011;31: Ahrens I, Peter K, Lip GYH, Bode C. Development and clinical applications of novel oral anticoagulants. Part II. Drugs under clinical investigation. Discovery Medicine 2012;13: Solomon DH, Van Houten L, Glynn RJ, et al. Academic detailing to improve use of broad-spectrum antibiotics at an academic medical center. Arch Intern Med 2001;161: O Brien T, Oxman AD, Davis DA, et al. Educational outreach visits: effects on professional practice and health care outcomes. Cochrane Database Syst Rev ;(4):CD Davis D, Galbraith R. Continuing medical education effect on practice performance effectiveness of continuing medical education: American College of Chest Physicians evidence-based educational guidelines. CHEST 2009:135: 42S 48S. 12. Mazmanian PE, Davis DA. Continuing medical education and the physician as a learner: guide to the evidence. JAMA 2002;288: Green LA, Gorenflo DW, Wyszewianski L. Validating an instrument for selecting interventions to change physician practice patterns: a Michigan Consortium for Family Practice Research study. J Fam Pract 2002;51: Soumerai SB, Avorn J. Principles of educational outreach ( academic detailing ) to improve clinical decision making. JAMA 1990;263: Kucher N, Koos S, Quiroz R. Electronic alerts to prevent venous thromboembolism among hospitalized patients. N Engl J Med 2005;352: Piazza G, Rosenbaum EJ, Pendergast W. Physician alerts to prevent symptomatic venous thromboembolism in hospitalized patients. Circulation 2009;119: Copyright 2013 by Turner White Communications Inc., Wayne, PA. All rights reserved. 454 JCOM October 2013 Vol. 20, No. 10