Use of Physician Extenders in Surgical Pathology Practice

Surgical pathology remains the area of service to which many pathologists devote the majority of their time.1 Despite increases in surgical pathology service time for many pathologists during the last several years, there is continuing pressure to improve pathology services (eg, decreasing turnaround times and lowering costs) from other providers and health care organizations in managed care systems. Workforce modifications have been one response to managed care pressures in many specialty practices.2–6 However, despite the critical role of workforce data in accurate assessment of staffing needs, evaluation and improvement of organizational efficiency, and assessment of quality of care and performance improvement, there is currently a lack of data on the health care workforce, including the pathology workforce.1,7,8 

For many years, clinicians in both primary and specialty care have used physician extenders (eg, physician assistants and nurse practitioners) for the performance of certain duties not requiring the expertise of a physician.9 Published studies examining clinical physician extenders have documented multiple advantages related to their use, including a positive impact on productivity and particular measures of quality, such as patient length of stay.4,10–12 Similarly, pathologists have used physician extenders for assistance with surgical pathology duties, primarily gross examination of surgical specimens, for decades. A growing proportion of these extenders are pathologists' assistants (Path Assts), laboratory personnel specifically trained to assist pathologists in anatomic pathology,13 either by graduating from a Path Asst training program accredited by the National Accrediting Agency for Clinical Laboratory Sciences or by on-the-job training. The American Association of Pathologists' Assistants offers a Fellowship examination for Path Assts who have met certain educational and training requirements.14 In lieu of an established national certification process for Path Assts, passage of this examination is used by some employers as a standard of competency. Previously published studies examining physician extenders in pathology have focused on Path Assts.13,15–17 However, pathologists also use other types of laboratory personnel (eg, medical technologists) as extenders in surgical pathology. The formal training programs for these various laboratory professionals generally do not include training in surgical specimen gross examination, the major task pathologists want extenders to perform in surgical pathology. Therefore, these types of laboratory personnel become extenders primarily through on-the-job cross-training in surgical pathology. Currently, laboratory accreditation by the Joint Commission on Accreditation of Healthcare Organizations and College of American Pathologists (CAP) requires compliance with multiple standards regarding the competency of laboratory staff.18 

The purpose of this study was to gather currently lacking descriptive data regarding the use of laboratory personnel as physician extenders in surgical pathology for application in future studies examining the impact of physician extenders on laboratory efficiency and quality of care.

Approval from the Allegheny General Hospital (Pittsburgh, Pa) institutional review board was obtained prior to initiation of this study. Distribution of questionnaires took place during the year 2001, and data entry and analysis were performed from January through June, 2002.

Two samples were used for this study; the first was a random, geographically balanced, cross-sectional sample of US pathologists practicing anatomic pathology (n = 968), and the second was a convenience sample of pathologists identified as directors of anatomic and/or surgical pathology at US academic centers (n = 76). The list of names for the first sample was purchased from a commercial provider of physician names and specialties associated with the American Medical Association. The list of names for the second sample was obtained from a directory of US pathology training programs.19 

The survey instrument used for the random sample of anatomic pathologists was a self-administered, mailed, voluntary, anonymous questionnaire designed by the study investigators to obtain descriptive information relating to the respondents' current use of and the scope of responsibilities of physician (pathologists and pathology residents) and multiple physician extender personnel types in surgical pathology. The physician extender personnel types specifically named in the questionnaire items were Path Assts, histotechnicians, histotechnologists, medical technologists, medical laboratory technicians, and cytotechnologists. Respondents were able to write in alternative personnel types, if applicable, for each of the items.

The survey instrument used for the random US sample of anatomic pathologists was composed of 13 items. The first 5 items related to the respondents' current practice characteristics (practice type, number of pathologists, annual number of surgical specimens, and number of beds in his or her primary practice setting). Items 6 and 7 asked the respondents to indicate all personnel types responsible for performing gross examination of surgical specimens in their practices. Item 8 asked respondents using physician extenders to indicate the reason(s) that best described why their practices used physician extenders. Item 9 asked respondents to indicate for all personnel types performing gross examinations (1) the number employed, (2) the full-time equivalent status, and (3) the average percentage of daily time spent performing gross examinations. Item 10 asked respondents to indicate what types of specimens each personnel type examined in their laboratories (biopsies, small routine specimens, large resections). Item 11 asked respondents using Path Assts to indicate the training background (program training or on-the-job training) of the Path Asst(s) they used. Item 12 asked respondents using physician extenders to indicate additional tasks extenders performed in their laboratories (frozen section preparation, specimen photography, etc). Item 13 asked respondents to indicate, on a 4-point scale, their level of satisfaction with the gross examination performance of each of the physician and physician extender personnel types involved in gross examination of surgical specimens in their laboratories (1, highly satisfied; 4, highly dissatisfied)

The survey instrument used for the directors' sample was similar to that used for the random sample of anatomic pathologists. Differences were that the directors' questionnaire did not ask for the full-time equivalent status and average percentage of daily time spent performing gross examinations of physician extenders; the 2 alternative items that were asked instead were (1) length of time the practice has used physician extenders in surgical pathology, and (2) whether the director favors national certification for Path Assts. The sample of directors of anatomic and/or surgical pathology was initially approached for telephone administration of the same questionnaire used for the random sample; however, owing to an initial poor response rate to the telephone questionnaire and the greater convenience of a self-administered questionnaire, the questionnaire was shortened (as described above), and the directors' survey ultimately was administered to the majority of this sample as a voluntary, anonymous, mailed questionnaire.

A modification of the Dillman total survey design approach was used to maximize the survey response rate,20 with follow-up postcard reminders sent approximately 3 weeks after the initial mailing to nonresponders.

All quantitative analyses were performed using the statistical software program SPSS (version 9.0; SPSS Inc, Chicago, Ill). Initially, descriptive statistics were used to describe the data in terms of proportions and central measures of tendency and dispersion.

Internal validity of questionnaire responses was examined by testing for a positive linear relationship between the self-reported number of full-time pathologists and the numbers of annual surgical specimens and beds in the primary practice setting using the correlation coefficient. Additionally, it was expected that the average annual number of surgical specimens would vary depending on practice type, with solo practices expected to process the smallest numbers of specimens and reference laboratories the largest. Differences in reported mean numbers of annual surgical specimens by practice type were examined using the 1-way analysis of variance (ANOVA).

Staffing ratios were calculated as follows: (Self-Reported Annual No. of Surgical Specimens Processed) × (%Daily Time Personnel Type Spent Performing Gross Examinations)/(Self-Reported No. of Full-Time Equivalent Spending at Least Part of Their Daily Time Performing Gross Examinations). These calculations were performed by personnel type (pathologist, Path Assts, etc).

Relationships between responses to other relevant questionnaire items were examined using a Fisher exact test, the χ² test, or correlation analyses. Differences between mean satisfaction values were examined using the 1-way ANOVA. Statistical significance was assumed at P ≤ .05.

The overall response rate was 22% (n = 225), with 193 responses from the random sample (20% response rate) and 32 responses from the director sample (42% response rate). For several of the items, small percentages of questionnaires contained missing data. Analyses involving less than the total number of responses from each of the samples (ie, 193 and 32) due to missing data are specified below, along with the actual numbers of responses analyzed.

The geographic distribution of all respondents is summarized in Figure 1, which shows that the majority of respondents were from southern and western regions of the United States.

The practice-type distribution of the random sample respondents is summarized in Figure 2. This distribution was very similar to the practice-type distribution for US pathologists reported by a recent CAP Practice Characteristics Survey.21 Similar to pathologist respondents to the CAP survey, the majority of study respondents reported being in pathology group practices.

Correlation analyses were performed between the variable self-reported number of full-time pathologists and the 2 following variables: self-reported annual number of surgical specimens processed and self-reported total number of beds in the primary practice setting. These analyses included 170 US sample responses and 32 director responses, which were combined for the analysis. A statistically significant positive correlation was found between the number of full-time pathologists and the annual number of surgical specimens processed (r = +0.609, P = .01) and between the number of full-time pathologists and the total number of beds in the primary practice setting (r = +0.661, P = .01).

The distribution of the average annual number of surgical specimens processed in respondents' laboratories is summarized in Table 1 and Figure 3. These statistics were calculated using 170 responses from the US random sample, owing to missing data. Table 1 presents the descriptive statistics and the ANOVA results for the annual number of surgical specimens reported by the random US sample, examined by practice type. The mean number of annual surgical specimens was significantly lower for solo practices when compared to the mean for multispecialty group practices (P = .004) and private/commercial reference laboratories. Pathology group and hospital employee practice means were also statistically significantly lower than the mean for private/commercial reference laboratories (P = .02 and P = .009, respectively). Figure 3 compares the distributions of surgical specimen volumes between the 2 study samples, showing the relatively larger volumes processed by the directors' academic practices (P < .001).

The proportions of respondents reporting use of physician extenders for surgical specimen gross examination by personnel type is shown in Table 2. Of the random US sample of pathologists, 56% of all respondents reported using physician extenders in surgical pathology. Since a proportion of each of the samples' respondents reported using a combination of personnel types, the sum of each sample column is greater than 100%. Fourteen percent of the US random sample reported using non–Path Asst physician extenders alone (eg, only histotechnicians, only medical laboratory technicians, or some combination) for assistance with surgical gross examinations, compared to none of the respondents from the directors' sample. All respondents from the directors' sample who used non–Path Asst physician extenders for gross examinations described using these personnel in combination with Path Assts. For those practices from both samples reporting the use of Path Assts for surgical gross examination, there was no statistically significant difference in the numbers of program-trained Path Assts employed compared to the numbers of Path Assts trained on the job. Respondents from the directors' sample who reported using physician extenders reported using extenders in surgical pathology for an average of 7 years (range, 1 to >20 years).

Certain practice types were associated with patterns of physician extender use at frequencies greater than expected by chance alone. These associations are shown in Table 3.

From the US random sample responses, we calculated the mean numbers of personnel types employed full time in practices spending at least part of their time performing gross examinations and the mean proportions of daily time these personnel types were reported to spend performing gross examinations. Means were calculated only for the subgroup of respondents reporting use of the particular personnel type. These values are shown in Table 4, along with mean staffing ratios for each personnel type. A significant positive linear relationship existed only between the number of pathologists and the total number of Path Assts in practices (r = +0.286; P = .01).

Statistically significant associations were found between particular personnel types and the types of specimens they examined. Pathologists' assistants more frequently performed gross examination of large resections (P < .001), and histotechnicians/histotechnologists more frequently performed gross examination of biopsies and small routine specimens exclusively (P = .002). A small proportion of Path Assts from both samples (approximately 10%) were used to gross biopsies and small routine specimens exclusively.

Respondents from both samples who reported using physician extenders for gross examination frequently reported using extenders for additional tasks in surgical pathology (46% of the random sample respondents and 78% of the directors' sample respondents). The most frequently reported additional task by both groups of respondents was assisting with preparation of frozen sections. A minority of respondents from both samples who reported using Path Assts also reported using Path Assts to assist with autopsy prosection (30% from the random sample and 19% from the directors' sample).

Mean values for pathologist level of satisfaction with physician extender gross examination performance by personnel type are summarized in Figure 4 (1, highly satisfied; 2, somewhat satisfied; 3, somewhat dissatisfied; and 4, highly dissatisfied). A 1-way ANOVA revealed no statistically significant differences between these mean values.

Reported reasons for use of physician extenders of all types by both the US random and directors' samples are summarized in Table 5. The reason reported most frequently by respondents from both samples was optimization of use of the pathologist's time. A minority of pathologists from the US random sample cited cost as a factor impacting physician extender use. Consistent with their academic practice settings, directors frequently offered reasons germane to pathology residency training.

Of the 32 director responses, 52% (n = 17) were in favor of creation of a national certification process for Path Assts, 33% (n = 10) were not in favor, 12% (n = 4) were either unsure or did not care, and 4% (n = 1) gave no response.

Of the 17 respondents in favor, 14 cited an expected establishment and/or improvement in the quality of Path Asst training and performance as a reason for their support of national certification. Other reasons given for favorable opinions were (1) the establishment of formal recognition for the Path Asst profession, (2) the expected formal oversight of the Path Asst profession by pathologists and pathology national organizations, and (3) the establishment of a standardized criterion useful for hiring decisions.

Of the 10 respondents not in favor, 8 described a lack of necessity for a certification process. Other concerns raised by those not in favor were (1) a potential loss of ability for pathologists to train extenders on the job, (2) a potential impact of certification on Path Asst salaries, and (3) a philosophical disagreement with the use of Path Assts in pathology practice. Those respondents raising the latter concern were opposed to national certification because generally they were opposed to any programs or policies supporting or promoting the Path Asst profession.

This study provides the first descriptive data regarding use of physician extenders in surgical pathology practice, information that is useful for tracking the nature and extent of extender use over time, as well as for future studies examining the impact of staffing on laboratory processes and outcomes.

The majority of pathologists responding to our random sample survey worked in group practices and reported using physician extenders to assist with gross examination of surgical pathology specimens, as well as with other duties in surgical pathology, such as preparation of frozen sections. While many pathologists used laboratory personnel specifically trained to assist pathologists in anatomic pathology (ie, Path Assts), a significant number of pathologists used other types of laboratory personnel as physician extenders, including clinical laboratory personnel, such as medical technologists.

Current Clinical Laboratory Improvement Amendment (CLIA) regulations for performance of high-complexity testing (493.1489) specify training requirements for laboratory personnel performing gross examinations. Additionally, all gross examinations performed by nonpathologist individuals must be reviewed by a pathologist. However, pathologists using non–Path Asst extenders for supervised gross examination tended to limit examinations by these extender types to biopsy and small routine specimens exclusively. If pathologists delegated supervised gross examination of large and/or complex specimens to physician extenders, they tended to delegate to Path Assts. This pattern of delegation is consistent with the more formal training, in general, of Path Assts in gross examination of surgical specimens. The use of multiple personnel types with varied training backgrounds for gross examination of biopsy specimens and small, routine specimens suggests that many pathologists believe satisfactory supervised gross examination of these specimen types does not require formal specialized training. This view is further supported by our finding that pathologists were equally satisfied with the gross examination performance of all laboratory personnel types that were used as physician extenders.

Statistically significant associations were found between certain practice types and patterns of physician extender use (Table 3). While one can hypothesize that factors such as cost or skill mix might significantly impact the practical choice of personnel type in a particular practice, we did not explore the specific reasons underlying a pathologist's choice of personnel type. While it is clear that the pathologists in our sample used extenders in general to optimize their own use of time, examination of the various factors impacting the choice of particular personnel types in a given practice setting were beyond the scope of the present study and were not addressed with our survey. These factors presently are unclear and are worthy of further study.

To our knowledge, the annual numbers of surgical specimens examined to laboratory personnel staffing ratios presented in Table 4 represent the first attempt to describe surgical pathology gross examination practice by physician extenders. These staffing ratios are highly variable, most likely reflecting not only the self-reported nature of our data, but also the high level of complexity of workflow and staffing patterns in current surgical pathology practice. Despite the variability in all the calculated ratios, the ratios were consistently highest for Path Assts, supporting the idea that Path Assts, as a group, are the personnel type responsible for the majority of supervised gross examinations performed in laboratories using physician extenders for gross examination of surgical specimens. Our findings also support the idea that in some practices, non–Path Asst personnel are responsible for significant numbers of supervised gross examinations. The implications of this finding in the context of the current medical laboratory personnel shortage22 for essentially all of these non–Path Asst personnel types are unclear.

Based on our previous studies of Path Assts in academic pathology practice, we expected the use of physician extenders to differ between academic and private practices, both in the extent of use of extenders (relatively more use of extenders, specifically Path Asst extenders, in academic practice) and in their role in surgical pathology practice (supervisor/teacher in academic practice).17 Collection of responses from a second, targeted sample of academic pathologist directors of anatomic and/or surgical pathology confirmed these expected differences and provided data contrasting physician extender use in academic practice versus private group practice, which the random sample responses largely represent.

One potential limitation of this study is the relatively low response rate, particularly for the US random sample of pathologists, limiting the ability to generalize our findings to the general US population of pathology practices. However, the strong similarity in practice-type distribution between our random sample respondents and recent CAP practice surveys supports the representative nature of our responses in terms of practice type, a variable logically expected to significantly impact physician extender use. Another potential limitation is the self-report nature of the data. Our ability to demonstrate the appropriate logical quantitative relationships between particular questionnaire items (eg, reported number of practice pathologists and reported annual number of surgical specimens) does support the validity of the responses we received.

The small total number of opinions collected from the directors' sample regarding national certification of Path Assts precludes making any strong conclusions regarding the general feeling of pathologists on this issue. However, the fact that there was frequent repetition of 1 reason for certification (potential standardization of a minimum level of training and performance quality) and 1 reason against certification (it is not necessary for good practice) suggests that the concept of quality of care is an important issue for all pathologists considering the use of physician extenders in surgical pathology; however, they do not agree on the potential impact of physician extender use on quality of care. Previous health care workforce studies of nonphysician personnel have shown staffing affects efficiency and quality of care.23–35 These studies have shown that both staff numbers and staff mix are important in achieving optimum performance. Further studies are needed to explore the impact of physician extender staffing patterns on surgical pathology laboratory efficiency and quality.