Physician Satisfaction With Clinical Laboratory Services: A College of American Pathologists Q-Probes Study of 138 Institutions

A high-quality organization meets customers' needs. Clinical laboratory managers often assume they know what customers want (accuracy, precision, speed, economy, etc) and set out to directly measure laboratory performance in each specific area. Another approach to measuring quality is to assess customer satisfaction with services without making any assumptions about the relative importance of specific functions.

Each approach to assessing quality has advantages. Direct assessment of accuracy, precision, and turnaround time helps managers understand whether local performance is improving and how it compares to published norms. The quality of processes can also be examined in great detail: the turnaround time for troponin tests, for example, can be measured separately from the turnaround time for potassium tests. On the other hand, measurement of customer satisfaction brings customer preferences into the quality assessment process and corrects for mistaken assumptions about which particular aspects of service customers value most.

Today, assessing customer satisfaction with laboratory services is considered an important component of a laboratory quality assurance program and is required for accreditation by the College of American Pathologists (CAP) and The Joint Commission on Accreditation for Healthcare Organizations.1,2 Physicians are one of the primary customers of laboratory services and obtaining their feedback provides laboratory managers with opportunities to identify areas for improvement.

The CAP's Q-Probes laboratory quality improvement program, which has produced numerous publications defining performance benchmarks in pathology and laboratory medicine,3 has been used to provide a standardized survey tool for assessing customer satisfaction with laboratory services.4–11 Two previous studies, performed in 1999 and 2002, reported an overall physician satisfaction score for clinical laboratory services of 4.0 (on a scale of 1 [poor] to 5 [excellent]) for the median laboratory, with the greatest satisfaction in the areas of staff courtesy, critical value notification, and quality of results, and the least satisfaction in the categories of turnaround times.4,10 Individual laboratories in the United States are required to repeat their assessment of customer satisfaction on a periodic basis.1,2 Periodic reassessment of group performance provides an opportunity to identify and characterize industry changes in service performance and customer attitudes. This study evaluated physician overall satisfaction and specific satisfaction of physicians with 15 aspects of clinical laboratory services.

Participants in the voluntary subscription Q-Probes program of the CAP collected data for this study during the first quarter of 2007. The study was composed of 2 parts: a standardized satisfaction survey with instructions for how to administer the survey to physicians and a general questionnaire about local practices that laboratory management completed. The survey form provided the mechanism for physicians to rate their overall satisfaction and their satisfaction with 15 specific aspects of clinical laboratory services. We entitled one aspect of service “quality/ reliability of results” to encompass the accuracy and precision of test results. Other aspects of service related to quality included turnaround time, accessibility of management, and other elements shown in Table 1. Evaluations used a scale of 1 to 5 with the following ranges: 1, poor; 2, below average; 3, average; 4, good; and 5, excellent. Physicians were asked to indicate which one of the 15 aspects of service was the most important to them. In addition, they were asked to respond to the question, “Would you recommend this laboratory to another physician?” with a yes or no answer. Laboratory participants distributed up to 300 surveys to physician customers. Data from all surveys that were returned within 4 weeks or the first 50 surveys returned, whichever occurred first, were compiled by laboratories onto a summary input document. A second input document represented a general questionnaire regarding laboratory characteristics and was completed by laboratory personnel to help provide comparison of characteristics between laboratories.

Statistical analysis of all data was performed by a CAP statistician (M.K.W.). Participating laboratories received individualized performance reports as well as a detailed data analysis and critique of the study results. Three composite satisfaction measures were used.

1. Overall satisfaction score: (No. of Excellent Ratings × 5) + (No. of Good Ratings × 4) + (No. of Average Ratings × 3) + (No. of Below Average Ratings × 2) + (No. of Poor Ratings × 1) for Overall Satisfaction/Total No. of Ratings (1–5) for Overall Satisfaction With Laboratory Services.

2. Percentage of excellent or good ratings: (No. of Excellent or Good Ratings for Specific Laboratory Service Category × 100)/ Total No. of Ratings (1–5) for Specific Laboratory Service Category.

3. Percentage of below average or poor ratings: (No. of Below Average or Poor Ratings for Specific Laboratory Service Category × 100)/Total No. of Ratings (1–5) for Specific Laboratory Service Category.

Institutional demographics and practice characteristics were evaluated for association with the 3 satisfaction measures. If a participant failed to answer a question for any of the demographics or practice characteristics listed, that participant's data were excluded from the database for that question only. Individual associations between mean overall satisfaction score and practice variables were investigated by using Kruskal-Wallis tests for discrete-valued independent variables and regression analysis for continuous-valued independent variables. Variables with significant associations (P < .100) were then introduced into a multivariable stepwise regression model. All variables remaining were significantly associated at the .05 significance level.

One hundred thirty-eight institutions participated in this study and submitted a total of 4329 physician surveys. Most participants were from the United States (97.1%), with the remainder from Australia (2), Singapore (1), and Spain (1). An average of 31.4 surveys were submitted per institution (median, 31), with a range of 3 to 50 surveys per institution. Response rate information was available for 123 laboratories. The mean and median response rates were 25.5% and 19.6%, respectively.

Approximately 34% of participating institutions were teaching hospitals and approximately 26% had a pathology residency program. Within the past 2 years, the CAP had inspected 85.7% of the laboratories, while 27% of laboratories were inspected by The Joint Commission. Institutional demographics and other characteristics are presented in Tables 2 and 3.

Table 1 displays the aggregate survey responses for the 15 survey categories. The average of the overall satisfaction ratings provided by the physicians was used as the laboratory's satisfaction indicator. The overall satisfaction scores ranged from 2.9 to 5.0, with a mean and median score of 4.1. In addition to this overall satisfaction score, a physician satisfaction score, based on the individual ratings of the 15 service categories, was calculated for each laboratory. This calculated satisfaction score had a strong correlation (r = 0.91) to the overall satisfaction ratings provided by the participating physicians, with a statistically significant positive association (P < .001) between the 2 satisfaction scores.

Respondents to the physician satisfaction survey indicated if they would recommend the clinical laboratory to another physician. All physicians surveyed from 57 of the 138 participating laboratories indicated that they would recommend that laboratory to another physician. The remaining 81 laboratories (58.7%) received at least one response in the category “would not recommend” the laboratory. Within this group, the percentage of physicians recommending the laboratory ranged from 60% to 98%. In the aggregate, 93.5% of physicians would recommend the laboratory (n = 4039). Table 4 displays the distributions of the overall satisfaction score and percentage of laboratories recommended by physicians. There was a strong association between the percentage of favorable physician recommendation of laboratories and the overall satisfaction score (P < .001).

All laboratory service categories except for esoteric test turnaround time had median percentage values of excellent/good ratings between 75.0% and 89.9%. Quality/reliability of laboratory results (analytical quality of results) and courtesy of laboratory staff had the highest median values (89.9%). Accessibility to laboratory staff, manager, and pathologist, and laboratory management responsiveness also had high median percentage values of excellent/ good ratings (range, 82.6%–87.6%). Of the 5 service categories that received the lowest median values for percentage of excellent/good ratings (combined scores of 4 and 5), 4 of these related to turnaround times for inpatient stat, outpatient stat, routine, and esoteric tests. These data are listed in Table 5.

The percentage of below average/poor ratings (combined scores of 1 and 2) for each laboratory service category was also calculated for each laboratory. Turnaround time for esoteric tests (tests sent out or not performed daily) had the highest median value (9.2%) for percentage of below average/poor ratings. Of the 5 service categories that received the highest percentage of below average/ poor ratings (combined scores of 1 and 2), 3 of these related to test turnaround times for inpatient stat, outpatient stat, and esoteric tests. Table 6 lists the data distribution.

Using a multivariable stepwise regression model, we evaluated participant demographics, percentage of inpatient/outpatient laboratory work, performance improvement activities, customer support services, and customer satisfaction practices to identify associations with overall mean satisfaction score (Tables 2 and 3). Only 2 laboratory characteristics were associated with satisfaction data. A higher mean overall satisfaction score was significantly associated with institutions reporting a higher percentage of billable tests performed in the combined categories of service for outpatients, nonpatients, and emergency department patients and in private institutions (nonprofit, profit, and independent laboratories) (Table 7).

Respondents to the physician satisfaction survey indicated which laboratory service category was most important to them. Almost one third selected quality/reliability of results as the most important category. This area of service also received the highest aggregate excellent/good ratings (combined scores of 4 and 5, 88.5%) and the lowest aggregate below average/poor ratings (combined scores of 1 and 2, 1.7%). The aggregate data for the findings on the most important category of service are listed in Table 8.

There are many definitions of quality. The Institute of Medicine, for example, has defined 6 elements of quality (patient safety, effectiveness, patient centeredness, timeliness, efficiency, equity) that it believes encompass important dimensions for medical care.12 In this investigation, we equated quality of clinical laboratory services with the degree to which a laboratory satisfies its primary customer—the physicians who order laboratory tests. In taking this approach, we assume that physician satisfaction with laboratory services will reasonably reflect the ability of a laboratory to meet the needs of patients, nonphysician caregivers, and society. We acknowledge, however, that any satisfaction instrument that surveys only one stakeholder group will provide an imperfect measure of satisfaction, and this limitation should be kept in mind.

In our investigation of 4329 physicians served by the 138 participating laboratories, we found a high level of overall satisfaction with laboratory services. Seventy-five percent of laboratories received mean satisfaction scores of 4.0 or higher (out of a possible score of 5) and a median overall performance score of 4.1. In similar studies of 375 institutions in 1999 and 153 institutions in 2002, 50% of the laboratories surveyed received median scores of 4.0 or higher, with 75% scoring 3.8 or higher.4,10 These scores are lower than the median satisfaction score of 4.4 reported in a study of customer satisfaction with anatomic pathology services.5 In the current study, surveys from half of the participating laboratories reported that 96% to 100% of physicians would recommend the laboratory to other physicians. This type of customer assessment, sometimes referred to as a net promoter score, has been found in some businesses to be a better measure of customer loyalty and for identifying problems; it better correlates with a company's revenue growth than do traditional customer satisfaction questionnaires.13,14 

Quality and/or reliability of results was listed as the most important category of laboratory services by 31.7% of physicians, more than twice as many physicians as those choosing the second most important category, routine test turnaround time (14.8%). This category was also the one with the highest level of physician satisfaction, with 89.9% of responses indicating a good or excellent evaluation for the median laboratory. These findings are comparable to the findings of the 2002 physician survey.10 It is unlikely that 68% of physicians—those who chose categories other than quality/reliability of results as the most important—do not value this category of service. Rather, quality/reliability of results has become such a reliable component of service that it is assumed to be an integral part of it. This allows focus on other service issues. This perspective may also explain the findings in a study of nursing satisfaction with clinical laboratory services, with accuracy of test results generating the highest level of satisfaction, despite the fact that it was listed as the second most important service category after stat test turnaround times.11 

Test turnaround times represented 3 of the top 5 categories listed by physicians as the most important (see Table 8); however, they were on the bottom for level of satisfaction (see Table 5). Comparable findings were reported in 2 previous physician and nursing satisfaction studies.10,11 In a previous study of satisfaction with anatomic pathology services, timeliness of reporting was also the service category of least satisfaction.5 Investment of resources for analyzing and improving preanalytic, analytic, and postanalytic processes to improve test turnaround times would clearly be the best focus for most laboratories to improve customer satisfaction. In particular, esoteric test turnaround time was the item with the lowest level of physician satisfaction, in both this study and the 2002 survey.10 This is undoubtedly related to the frequency with which these tests are performed, inherently long testing times for some tests (eg, cytogenetics), or to the fact that these tests are being sent to a reference laboratory. To improve physician tolerance and satisfaction, some effort with regard to physician education and to managing expectations may be as valuable as working with the testing or send-out processes.

Laboratories with a higher percentage of testing from sources other than inpatient services (outpatient, nonpatient, emergency department) tended to have a higher level of physician satisfaction. This is most likely because of the higher expectation for service in the competitive “outreach” environment to maintain customers that have choices in laboratory service providers. Private institutions (nonprofit, profit, and independent laboratories) also tended to have a higher level of physician satisfaction, with university and public/governmental institutions having the lowest level of satisfaction with their laboratory services. Possible explanations for these findings include organization differences in available resources, management philosophy regarding the importance of service compared with research and education, or different physician expectations among customers of different types of laboratories. This study was not designed to distinguish between these possibilities; further research is required.

Interestingly, we were unable to identify significant associations with physician satisfaction with respect to the laboratory practices of monitoring turnaround times, corrected reports, critical value notification, lost specimens, pending lists, customer complaints, telephone response times, the presence of dedicated customer support staff, or customer satisfaction training. Our inability to find an association may be due to any of several causes: (1) this study had low statistical power to detect small associations, (2) these activities may not have any impact on customer satisfaction, (3) detailed in-laboratory monitoring of operations may not be followed up with corrective action, or (4) facilities with poorer-than-average performance may have been more likely than average to monitor quality of their operations. More research is needed to help distinguish between these potential explanations.

Several limitations of this study should be recognized. First, as mentioned, a survey of physician satisfaction may imperfectly reflect the views of patients and other stakeholders. Second, although data collection was standardized among participants, only the first 50 responses were tabulated per site and it is possible that this approach introduced bias towards “early responders,” particularly among larger institutions where multiple responses were received. Finally, we studied a convenience sample of Q-Probes subscribers that may not be representative of the laboratory community as a whole. Nevertheless, the principal findings of this study are likely to be robust: (1) laboratories participating in this study have a high level of overall physician customer satisfaction and loyalty and (2) persistently lower levels of satisfaction with test turnaround times offer a significant opportunity in many laboratories for service improvement.