ORIGINAL ARTICLES. Information Primary Care Physicians Want to Receive About

Transcription

1 ORIGINAL ARTICLES Information Primary Care Physicians Want to Receive About Their Hospitalized Patients Kenneth J. Smith, MD, MS; Sunday Clark, ScD, MPH; Wishwa N. Kapoor, MD, MPH; Steven M. Handler, MD, PhD BACKGROUND AND OBJECTIVES: Communication between physicians caring for hospitalized patients and those patients primary care providers (PCPs) is often suboptimal, which can lead to diminished health care quality and safety. It is unclear what hospital information PCPs would find most valuable in their patients continuing care, as is how and when they would prefer to receive such information. METHODS: Using the modified Delphi survey methodology, we developed a consensus list of information items PCPs want to receive about their hospitalized patients, using general internists and family physicians considered experts in primary care. Panelists rated items on a 5-point Likert scale signifying their level of agreement with the information s importance and with the information communication mode. Consensus agreement or disagreement was determined using 95% confidence intervals. RESULTS: Twelve physicians (five family physicians, seven general internists), averaging 19.6 years of primary care experience, participated in Delphi round 1; 41.6% (37 of 89) of the items were accepted by consensus, one item was rejected (receiving daily progress notes), and the remaining 51 items were equivocal. In round 2, nine physician panelists participated (four family physicians, five general internists), and six additional items were accepted. They generally preferred notification at the patient s first hospital interaction and at discharge. No consensus was found regarding communication mode; was most favored. CONCLUSIONS: We found broad areas of consensus regarding information PCPs wish to receive about their hospitalized patients that are generally consistent with previous surveys. Our findings also suggest that physicians are becoming more comfortable with patient-related electronic communications. (Fam Med 2012;44(6): ) Hospitalized patients are increasingly being cared for by physicians other than their primary care providers (PCPs). 1,2 However, communication between physicians caring for hospitalized patients and those patients PCPs is often suboptimal, which can lead to diminished health care quality and safety. 3 It is hoped that better communication between hospitalist physicians and PCPs both during and after a patient s hospitalization will improve quality of care and decrease medical errors. 4-6 However, a recent multicenter trial has shown that PCPs receipt of information regarding their hospitalized patients had no effect on patient outcomes, suggesting that The presence of any communication is not the same as the receipt of high-quality communication, and only the latter may be capable of improving outcomes. 7 The characteristics and effects of such high-quality information are, as yet, undefined. Numerous studies and reviews have made recommendations to improve transitions of care from the hospital to the ambulatory setting, mainly focusing on communicating with the patient s PCP, documenting changes in the patient s medication regimen occurring during the hospitalization, planning adequate care for the patient during the transition, and facilitating communication between physician and patient. 4,5, 8-11 These recommendations tend to center on the hospitalists or the patients perspective but not that of PCPs. 4,5, 8-11 Little work has been performed to elucidate what information PCPs might find useful during the hospitalization to facilitate their patient s continuing care. 12 From the Department of Medicine (all) and Department of Biomedical Informatics (Dr Handler), University of Pittsburgh, Pittsburgh, PA. FAMILY MEDICINE VOL. 44, NO. 6 JUNE

2 Additionally, despite the availability of newer communication modalities, such as and Web-based tools, little is known about communication preferences of PCPs when communicating with hospitalists about their hospitalized patients. 12 Attempts to define high-quality communication should include the input of PCPs to guide the efficient transmittal and use of hospital information. Since there is an incomplete state of knowledge regarding PCP communication preferences, we used an Internet-based two-round modified Delphi survey to develop a consensus list of the information items PCPs want to receive about their hospitalized patients and how and when they prefer to receive this information. Methods For the purposes of our study, we drew candidate information items from the published literature, focusing on (1) communication between the hospital and the PCP and between the physician and patient, (2) changes in medication regimens during the hospitalization, and (3) care planning during transitions, concentrating on those data elements that PCPs might consider relevant in their patients post-hospital care. 4, 5,8-11 We also used published data regarding communication preferences of PCPs who have worked with hospitalists to develop candidate items. 12 Our intent was to include a broad range of information items that PCPs might find valuable in the continuing care of their patients, from which our PCP expert panel could choose the most valuable items. Candidate items are shown in the survey (available from author on request). Items were organized into categories including (1) emergency department evaluation, (2) admission-related information (the first 24 hours of a hospital stay), (3) interim information (>24 hours of a hospital stay), (4) any acute change in condition (such as cardiac arrest, respiratory failure, or acute mental status change), and (5) hospital discharge. We also asked the PCP expert panel for their preferences regarding communication modality (ie, , pager, phone, letter, direct communication to office staff). To obtain a consensus on the items PCPs might find most useful regarding their hospitalized patients, we used an Internet-based, two-round modified Delphi survey of PCPs who are considered experts in their respective fields of family medicine and general internal medicine. The Delphi methodology is a structured group interaction process, directed in rounds of opinion collection and feedback. 13,14 Research suggests that a modified Internetbased Delphi consensus method can provide accurate and reliable assessments of clinical or informational parameters by consulting a panel of experts and subsequently accepting the group consensus as the best estimate of the answer to a particular question. 15,16 This methodology differs from the Delphi process developed by the RAND Corporation, where face-to-face meetings are used to achieve consensus; 13,14 however, the modified Delphi allows easier, geographically unconstrained contact between experts. 15 Survey rounds were completed confidentially, allowing participants to present and react to ideas without being biased by knowing the other participants identities. PCPs chosen for the panel were either family physicians or general internists, recommended as experts in primary care, with the aim of having a physician panel representing various US regions, as well as both community and academic primary care practices. The goal was to have a similar number of participants from family medicine and internal medicine. Family physicians were recommended by physician members of the University of Pittsburgh Department of Family Medicine; general internists were recommended by general internal medicine division chiefs at several university programs. We invited 16 family physicians and 14 general internists to participate. Panelists received no compensation. Panelists were invited to participate via , with a maximum of three attempts to enlist their participation. Those who agreed were sent, via , the link to the survey website. The survey consisted of all the candidate information items; participants were asked to rate each item on a 5-point Likert scale (from 1=strongly disagree to 5=strongly agree) based on their level of agreement with the information item s importance in a patient s continuing care. We also asked them to use the same 5-point Likert scale to rate their agreement with statements regarding the communication modality they prefer for information on their hospitalized patients. During the first round, participants were also given the option to suggest additional items; if two or more panelists suggested the same item be added, it would be included in the second round of the survey. 15,17 The study was completed in March and April Once round 1 of the Delphi was completed, results were tabulated and 95% confidence intervals (CIs) calculated for each item. Similar to previous studies, an item whose lower 95% confidence limit was >4.0 (indicating consensus agreement) was denoted as accepted, and items with an upper 95% confidence limit <3.0 (indicating consensus disagreement) were rejected. 15,17 Remaining items that were neither accepted nor rejected were denoted as equivocal, and were included in round 2. In the Delphi second round, panelists received their own previous item responses and item average scores for all panelists. After completion, round 2 scores and 95% CIs were calculated, and additional items were accepted or rejected by consensus based on the criteria above. STATA version 11.0 (StataCorp, College Station, TX) was used for statistical analyses. Based on our prior work with Delphi panels, 15,18 we felt it would be difficult to sustain interest among our physician panelists and retain them in sufficient numbers beyond two Delphi rounds. This study was 426 JUNE 2012 VOL. 44, NO. 6 FAMILY MEDICINE

3 reviewed and considered exempt by the University of Pittsburgh Institutional Review Board, and informed consent was not necessary for study participation. Results Twelve physicians (five family physicians, seven general internists) participated in round 1 of the Delphi panel; five (42%) practice in community settings (the remainder practicing at urban, academic centers), and nine (75%) had their patients admitted regularly to a hospitalist service (of the remaining PCPs, two cared for some hospitalized patients, and one had patients admitted to other non-hospitalist services). They had an average of 19.6 years (standard deviation=8.3; range=8 30) of primary care practice experience. In round 1, based on the criteria outlined above, 41.6% (37 of 89) of the information items were accepted by consensus, one item was rejected (receiving daily progress notes), and the remaining 51 items were equivocal. No additional information items were suggested by two or more panelists. Nine of the PCPs who participated in round 1 (four family physicians, five general internists) participated in round 2, which consisted of the 51 equivocal items. In this round, 11.8% (6 of 51) additional items were accepted by consensus, with no consensus for acceptance or rejection found for the remaining items. All accepted items are shown in Table 1. As shown in Table 1, the panel generally preferred notification at the patient s first hospital interaction (eg, in the emergency department) and at discharge, without much contact in between, except for notification in the event of medical crises, such as cardiac arrest or respiratory failure. During the preadmission and admission phases, PCPs preferred to receive multiple data elements pertaining to the emergency department visit (physician documentation, laboratories, radiology, and medications), notification of admitting diagnoses, and consultant s evaluations. At discharge, PCPs preferred to receive a brief description of the hospital course (rather than an extensive one), discharge medication and medication reconciliation data, key findings and test results during the hospitalization, a listing of pending laboratories and other tests, and follow-up plans, among other data elements. After discharge, PCPs preferred to receive the results of pending laboratory, radiology, and diagnostic studies. PCPs came to no consensus regarding their preferred mode of receiving hospital information (Table 2). was most favored, with fax being the next most favored mode; regular mail was least favored. Trends against consensus acceptance or rejection were observed in round 2 compared to round 1. Discussion We found that PCPs prefer most communication regarding their hospitalized patients at the beginning and the end of a patient s hospitalization and that the information be concise and center on key findings, medications, and follow-up plans. During a patient s hospitalization, PCPs only wanted to be notified about medical crises; they did not want to receive daily hospital progress notes. While was the most favored mode of communication, the panel did not come to a consensus on how they would prefer to receive communications about their hospitalized patients. Our findings are consistent with published recommendations for improving care transitions at hospital discharge. 4,5 These recommendations include communicating the presenting problem, key findings and test results, the final diagnoses, discharge medications and changes in medications, follow-up plans, and pending tests on the day of discharge, followed by a more detailed discharge summary. The recommended discharge summary should be available within 1 week and contain additional information: a brief hospital course, functional status at discharge, consultant recommendations, documentation of patient education and understanding, and anticipated problems and suggested interventions. Our findings are also consistent, for the most part, with a 1998 survey of PCPs, where more than half of PCPs said they would prefer to communicate with hospitalists at admission, at discharge, at the time of a major change in status (eg, a transfer to ICU or a change in resuscitation status). 12 However, our results did not support notification at the time of major procedures, such as coronary or operative interventions, which was favored in that survey. 12 Our panel favored receiving this information at discharge, as with most other information pertaining to a patient s hospital stay. The prior survey indicated that PCPs preferred to receive information via telephone rather than ; 12 in our survey, was most preferred, although no consensus was reached. We expect that e- mail will continue to become more acceptable with time, as might other electronic means of communication, such as web-based communication systems. Our findings, however, could certainly be biased in favor of given the -based recruitment process we used for our panel. Also in keeping with prior surveys, 19,20 PCPs are interested in receiving information about their patients emergency department visits, an area where information transmittal to PCPs often falls short. 21,22 In our survey, we made no distinction between emergency department visits that resulted in hospitalization and those that did not (survey available from authors upon request). In one study, most physicians felt that emergency department visit information was essential in the continuing care of their patients. 20 Another study found that use of a Web-based standardized communication system between an emergency department and family physicians led to more frequent and more useful information exchange. 23 Efforts to improve hospital communication with PCPs should consider including FAMILY MEDICINE VOL. 44, NO. 6 JUNE

4 Pre-admission Status Table 1: Information Primary Care Physicians Favored Receiving From Hospitals Mean 95% CI Be notified of the patient s Emergency Department (ED) visit Receive physician documentation from the ED visit Receive consultant physician evaluation Receive ED medications discharged on Receive ED laboratory study results and pending results Receive ED radiology studies and reports Receive ED diagnostic studies (other) and reports Receive discharge instructions from the ED visit Admission-related information (data generated within the first 24 hours of admission) Be notified of the patient s admitting diagnosis(es) Receive consultant evaluation Change in status information (occurring anytime after admission) Be notified of medical crises (e.g., cardiac arrest, respiratory failure) Discharge-related information (defined as information sent on the day of discharge) Receive presenting problem Receive key findings and test results Receive discharge diagnosis(es) Receive discharge medications Receive comparison of discharge medications with admission medications (ie, medication reconciliation) Receive list of changes in dosage and/or frequency in previous prescribed medications and rationale for changes Receive list of medications started during the hospitalization including rationale for prescribing Receive list of medications discontinued during the hospitalization and rationale for discontinuing Receive pending laboratory and tests Receive follow-up plan Receive recommendations of subspecialty consultants Information included in/with a discharge summary Receive discharge diagnosis(es) Receive name/contact information of the discharging physician Receive brief hospital course Receive lab results Receive major procedures/treatments performed Receive results of procedures Receive responses to treatments Receive recommendations of subspecialty consultants Receive discharge medications Receive comparison of discharge medications with admission medications (continued on next page) 428 JUNE 2012 VOL. 44, NO. 6 FAMILY MEDICINE

5 Table 1: Continued Mean 95% CI Information included in/with a discharge summary (continued) Receive reasons for changes and indications for newly prescribed medications Receive drug allergies; adverse drug reactions Receive patient s functional status at discharge Receive patient s cognitive status at discharge Receive resuscitation/code status and other end-of-life issues Receive pending laboratories and tests Receive follow-up plan Receive future appointments, procedures, and laboratory studies Post discharge (defined as information generated following discharge) Receive results of pending laboratory studies Receive results of pending radiology studies Receive results of pending diagnostic studies Panel consensus was not reached until Round 2 of the Delphi survey. Means based on 5-point Likert scale with 1=strongly disagree and 5=strongly agree CI confidence interval Table 2: Primary Care Provider Preferences for Communication With the Hospital Communication Preferences Mean 95% CI Communications from the hospital should be via Communications from the hospital should be via fax Communications from the hospital should be via pager Communications from the hospital should be via phone Communications from the hospital should be via regular mail Communications from the hospital should be routed through my office staff Means based on 5-point Likert scale with 1=strongly disagree and 5=strongly agree CI confidence interval information about emergency department visits to PCPs, given the perceived need for and usefulness of these data. We did not include electronic medical records (EMRs) as a means of communication in our survey under the assumption that PCPs and hospitals are not necessarily part of the same health care system and, even if they are, they may not have compatible EMR systems or EMR at all. In hospitals, EMR systems could automatically compile information and send it, at the prescribed times and in clear and concise formats, to PCPs in the mode that they prefer to receive it, be it via EMR (if available), , or other means. We would not expect that PCPs preferred information would change based on mode of delivery. Automating the information sent from the hospital, rather than depending on hospital personnel to obtain and deliver it, might be preferred if automatically extracted data could be reliably and reproducibly sent to PCPs. This study has limitations. We chose PCPs to participate in the Delphi panel based on colleagues recommendations, not on objective criteria. We took a number of steps in an attempt to minimize bias related to participant selection. We included both family physicians and general internists so that a broader range of PCPs would be represented in our expert panel and for the same reasons assured that physicians in both community and academic practices and from different FAMILY MEDICINE VOL. 44, NO. 6 JUNE

6 geographic regions were included. We are limited, however, by low response rates, but low response rates among physicians in surveybased research are common. It is not clear how many experts are needed to support a Delphi procedure, although it is suggested that fewer experts are needed when their backgrounds are similar. 24 Another limitation, as mentioned above, is the electronic nature of our recruitment process and survey procedures. Finally, our Delphi procedure, unlike some other Delphi protocols, did not allow for direct face-to-face or telephone contact between panelists and hence may have hindered consensus being found on more survey items. We found broad areas of consensus regarding information PCPs wish to receive about their hospitalized patients. These findings are generally consistent with previous surveys. Our results also suggest that PCPs prefer to receive information about their hospitalized patients at the beginning and end of the hospital stay, with less communication during the hospital course. While we found no consensus on how PCPs prefer to receive information, limiting the conclusions that can be made, e- mail was most favored, suggesting that physicians may be becoming more comfortable with electronic communications regarding their patients. In addition, electronic communication tools have the potential to facilitate PCPs communication, allowing them to provide the information and perspective inherent in a patient continuity relationship and enhancing their patients care; future work investigating the results of heightened two-way communication between PCPs and hospitals is warranted. Finally, in an era where hospitalists appear to be a growing and enduring component of medical care, incorporating communication with and data from hospitalists into the workflow of office-based physicians and their trainees will be an important aspect of their practice, as well as an educational challenge at all levels of training. Whether modern communication systems can efficiently and effectively transmit information that PCPs find valuable, facilitate interaction between office- and hospital-based physicians, and, most importantly, improve patient care requires further directed study in these areas. ACKNOWLEDGMENTS: Funding for this project was received from the US Agency for Healthcare Research and Quality (R18 HS and R01HS018721). CORRESPONDING AUTHOR: Address correspondence to Dr Smith, University of Pittsburgh, Department of Medicine, 200 Meyran Ave, Suite 200, Pittsburgh, PA Fax: smithkj2@upmc.edu. References 1. Wachter RM, Goldman L. The emerging role of hospitalists in the American health care system. N Engl J Med 1996;335(7): Wachter RM, Goldman L. The hospitalist movement 5 years later. JAMA 2002;287(4): Kripalani S, LeFevre F, Phillips CO, Williams MV, Basaviah P, Baker DW. Deficits in communication and information transfer between hospital-based and primary care physicians: implications for patient safety and continuity of care. JAMA 2007;297(8): Halasyamani L, Kripalani S, Coleman E, et al. Transition of care for hospitalized elderly patients development of a discharge checklist for hospitalists. J Hosp Med 2006;1(6): Kripalani S, Jackson AT, Schnipper JL, Coleman EA. Promoting effective transitions of care at hospital discharge: a review of key issues for hospitalists. J Hosp Med 2007;2(5): Jack BW, Chetty VK, Anthony D, et al. A reengineered hospital discharge program to decrease rehospitalization: a randomized trial. Ann Intern Med 2009;150(3): Bell CM, Schnipper JL, Auerbach AD, et al. Association of communication between hospital-based physicians and primary care providers with patient outcomes. J Gen Intern Med 2009;24(3): Coleman EA, Mahoney E, Parry C. Assessing the quality of preparation for posthospital care from the patient s perspective: the care transitions measure. Med Care 2005;43(3): Coleman EA, Parry C, Chalmers S, Min SJ. The care transitions intervention: results of a randomized controlled trial. Arch Intern Med 2006;166(17): Goldman L, Pantilat SZ, Whitcomb WF. Passing the clinical baton: 6 principles to guide the hospitalist. Am J Med 2001;111(9B):36S-39S. 11. Roy CL, Poon EG, Karson AS, et al. Patient safety concerns arising from test results that return after hospital discharge. Ann Intern Med 2005;143(2): Pantilat SZ, Lindenauer PK, Katz PP, Wachter RM. Primary care physician attitudes regarding communication with hospitalists. Am J Med 2001;111(9B):15S-20S. 13. Brook RH, Chassin MR, Fink A, Solomon DH, Kosecoff J, Park RE. A method for the detailed assessment of the appropriateness of medical technologies. Int J Technol Assess Health Care 1986;2(1): Wyrwich KW, Fihn SD, Tierney WM, Kroenke K, Babu AN, Wolinsky FD. Clinically important changes in health-related quality of life for patients with chronic obstructive pulmonary disease: an expert consensus panel report. J Gen Intern Med 2003;18(3): Handler SM, Hanlon JT, Perera S, et al. Consensus list of signals to detect potential adverse drug reactions in nursing homes. J Am Geriatr Soc 2008;56(5): Jones J, Hunter D. Consensus methods for medical and health services research. BMJ 1995;311(7001): Lindblad CI, Hanlon JT, Gross CR, et al. Clinically important drug-disease interactions and their prevalence in older adults. Clin Ther 2006;28(8): Smith KJ, Zimmerman RK, Lin CJ, et al. Alternative strategies for adult pneumococcal polysaccharide vaccination: a cost-effectiveness analysis. Vaccine 2008;26(11): Choyce MQ, Maitra AK. Satisfaction with the accident and emergency department a postal survey of general practitioners views. J Accid Emerg Med 1996;13(4): Wass AR, Illingworth RN. What information do general practitioners want about accident and emergency patients? J Accid Emerg Med 1996;13(6): Dunnion ME, Kelly B. From the emergency department to home. J Clin Nurs 2005;14(6): Jansen JO, Grant IC. Communication with general practitioners after accident and emergency attendance: computer generated letters are often deficient. Emerg Med J 2003;20(3): Afilalo M, Lang E, Leger R, et al. Impact of a standardized communication system on continuity of care between family physicians and the emergency department. CJEM 2007;9(2): Hsu CC, Sandford BA. The Delphi technique: making sense of consensus. Pract Assess Res Eval Available at pareonline.net/pdfv12n10.pdf. 430 JUNE 2012 VOL. 44, NO. 6 FAMILY MEDICINE