Objective: This study evaluates emergency physicians' perceptions of the malpractice risk posed by
utilization of physician assistants (PAs) in emergency departments (EDs) in 2004 and again in 2009.
Methods: A 16-question survey was mailed to a random sample of 1,000 active members of the American College of Emergency Physicians in 2004 and again in 2009.
Results: In both 2004 and 2009, 70% of the emergency physicians did not believe that PAs, when properly supervised, are more likely to commit malpractice than any other clinician. In both surveys, 80% of the respondents did not believe PAs were more likely to be sued for malpractice. A significant negative correlation was found between perceived risk of malpractice by PAs and the number of years physicians had worked with PAs or had worked in emergency medicine. From 2004 to 2009, the number of respondents practicing with PAs increased by 26%, the number directly supervising a PA in the ED increased by 19%, the number who thought PAs decreased patient wait times increased by 13%, and the number who reported that PAs increased patient satisfaction increased by 10%.
Conclusion: As physicians gain clinical experience with PAs, their perceived risk of malpractice tends to decrease. These results may have implications for the utilization of PAs, particularly as EDs become more utilized for noncritical situations.
As the physician assistant (PA) profession evolves with time, more PAs have chosen to practice in specialty and subspecialty areas. In 1996, American Academy of Physician Assistants (AAPA) census data showed that 39.8% of PAs were practicing in family medicine.
1 By 2008, that percentage had declined to 25.9%.
2 Over a similar period of time, one medical specialty in which PAs practice, emergency medicine, has also realized an increase in patient utilization. Emergency department (ED) overcrowding has increased each year. A survey conducted by the US Department of Health and Human Services found that from 1996 to 2006, visits to EDs increased by 32%.
3 It has been suggested that overcrowding and increased wait times have led to compromises in patient care, an increased risk for patients leaving against medical advice, and poor patient satisfaction.
3-7
The utilization of fast tracks in the ED has been shown to decrease patient wait times and length of stay.8 In addition, increased use of and expanded roles for PAs and nurse practitioners (NPs) in the ED9 may allow physicians to be available for more critical cases. Physician assistants now treat approximately 10% of the total patient population that visits EDs each year.3
This degree of PA utilization may be influenced by many perceptions, one of which may be the effect of PA utilization on overall malpractice risk. For instance, PA utilization may be sacrificed because PAs are believed to increase malpractice risk for the supervising physician. However, it has been suggested that PAs decrease overall medical malpractice risk as determined by the frequency and amount of reported malpractice payments tracked by the National Practitioner Data Bank.10-12 Furthermore, evidence suggests that a decrease in patient wait times and effective communication, which increase patient satisfaction, may lead to a decrease in malpractice litigation.10,13,14 The objective of the current study was to determine emergency medicine physicians' perceptions of PA malpractice risk in the ED; the perceived effect of PAs on patient wait times and patient satisfaction; the factors that may influence the perceptions of PA malpractice risk; and if these perceptions have changed from 2004 to 2009.
METHODS
A 16-question survey was mailed to 1,000 active members of the American College of Emergency Physicians (ACEP) between December 2003 and January 2004, and again between December 2008 and January 2009. The samples were selected randomly by the ACEP in both years. The survey was developed in 2004 under the guidance of a pilot sample of 10 emergency medicine physicians and PAs, university research faculty, and officers of the Society of Emergency Medicine Physician Assistants (SEMPA). The survey and project were approved by the Institutional Review Board of Midwestern University, Glendale, Arizona, on both occasions.
The survey consisted of three sections: 1) emergency medicine physicians' perceptions of PA malpractice risk, utilization, and patient satisfaction; 2) factors that may influence perceived malpractice risk; and 3) general practice demographics of the participants, such as practice location and years of clinical experience. In the first section, physicians expressed their opinion on statements that utilized a four-point Likert scale that ranged from strongly agree (1) to strongly disagree (4). The "forced-choice" method was employed to avoid neutral or nonapplicable answers. In the second section, physicians rated four factors that may influence perceived PA malpractice risk. This was based on a similar four-point Likert scale that ranged from significantly decreases risk (1) to significantly increases risk (4).
Each Likert response was assigned a numerical value for data entry. The four factors that may influence PA malpractice risk were then ranked according to the mean Likert value. Descriptive statistics were calculated on each question's response (means, frequencies, and percentages). In addition, the two survey questions examining physicians' perceived malpractice risk (questions 1 and 2 in Table 2) by PAs were collapsed into a new variable entitled risk. A two-tailed Pearson correlation was conducted to examine the relationship between the variables such as years of emergency experience and years of experience working with emergency medicine PAs with the variable risk. The internal validity of the surveys was assessed using calculations of Cronbach α.
RESULTS
In 2004, 401 surveys were returned (40.1%), and in 2009, 323 were returned (32.3%). In 2004 and 2009, approximately 97% of the respondents were currently practicing emergency medicine, 87% were board-certified in emergency medicine, and the respondents had been practicing emergency medicine for an average of 14 years (Table 1). During the 5-year period, there was a 26% increase in the average number of years the physicians reported working with PAs in emergency medicine (6.1 versus 7.7 years) and a corresponding increase in the number of respondents who were either supervising physicians or agents for PAs in emergency medicine (64.7% versus 76.7%). In 2004 and 2009, the most common work settings in which PAs were employed were a combination of hospital emergency and hospital fast-track/urgent care (34%), hospital emergency department only (27%), or hospital fast-track/urgent care only (12%).
The percentage of physicians who disagreed or strongly disagreed that PAs are more likely than physicians to commit medical malpractice was 71.6% in 2004 and 67.9% in 2009. In 2004, 84.3% of the respondents, and in 2009, 81.8% of the physicians disagreed or strongly disagreed that PAs were more likely than physicians to be sued as a result of medical malpractice. Additionally, most of the respondents either agreed or strongly agreed that PAs in the ED decrease wait times for patients to be seen (85.2% in 2004; 91% in 2009). Likewise, 65.2% in 2004 and 75.4% in 2009 agreed or strongly agreed that PAs increase patient satisfaction (Table 2).
Factors that may influence malpractice risk were ranked according to the total number of respondents who chose significantly decreases risk. In both surveys, respondents reported that malpractice risk would decrease the most as the years of clinical experience by the PA increased (overall mean on Likert scale = 1.40). Completion of a postgraduate residency program was ranked second in order of importance to decrease malpractice risk (1.53), appropriate supervision of PAs by physicians ranked third (1.67), whereas a higher level of education by the PA was the least significant factor believed to decrease malpractice risk (1.83) (Table 3).
A negative correlation was found in 2004 and 2009 between physicians' perceived risk of malpractice by PAs and the number of years of experience working with PAs
(r = -0.23, P < .01 in 2004; r = -0.21, P < .001 in 2009),
or the number of years working in emergency medicine
(r = -0.21, P < .01 in 2004; r = -0.12, P < .05 in 2009). The Cronbach a was 0.69 in 2004 and 0.73 in 2009.