Objectives.—To establish the rates with which reference laboratories resolve inquiries telephoned to them from primary laboratories and to identify reference laboratory practices associated with higher rates of inquiry resolution.
Design and Participants.—For 2 months, or until 50 contacts had occurred, 545 primary laboratories participating in the College of American Pathologists Q-Probes laboratory quality improvement program prospectively documented and characterized telephone inquiries they made to a reference laboratory of their choice. Participants also cataloged their own laboratory's demographic and practice characteristics and their reference laboratory's customer service characteristics.
Main Outcome Measure.—Rates with which reference laboratories resolved telephone inquiries.
Results.—Participants characterized 11 031 (78.7%) of 14 017 telephone inquiries as resolved by the reference laboratories. Ranked according to inquiry resolution rates, primary laboratories in the 90th percentile characterized reference laboratories as resolving 100% of their inquiries; those in the 10th percentile characterized reference laboratories as resolving only 54.2% of their inquiries. The rate of resolved inquiries was significantly higher (P = .0047) for participants using reference laboratories with 24-hour customer service than it was for participants using reference laboratories with less than 24-hour service. Most primary laboratories (80.9%) chose to monitor 1 of 11 national reference laboratories; in this subset, median rates of inquiry resolution ranged from 90.2% to 55.0% (P < .0001), despite no significant variation in other measured customer service characteristics.
Conclusions.—Primary laboratories experience significant differences in the rates with which reference laboratories resolve telephone inquiries. The performance benchmark for reference laboratories is resolution of at least 90% of telephone inquiries from primary laboratory customers.
Medical laboratories that provide primary services to their medical communities refer less frequently ordered or complex tests to specialized, or reference, laboratories that are better suited to perform those tests. As a requirement for laboratory accreditation by the College of American Pathologists (CAP), primary service laboratories must have an effective mechanism for evaluating their reference laboratories.1 Particularly useful tools in evaluating performance include benchmarking and internal auditing of the quality of the goods and services delivered.2–5 To date, the medical literature contains no reports that compare the performance of reference laboratories.
Since 1989, the CAP Q-Probes Program has conducted multi-institutional studies that have established a broad range of performance benchmarks in laboratory medicine.6 This Q-Probes study sought to develop a model that primary service laboratories could use to quantify reference laboratories' service in the important area of response to telephone inquiries. We characterized the types of telephone inquiries that personnel in primary laboratories made to reference laboratories and measured the rate with which reference laboratory personnel resolved the inquiries. We chose to benchmark resolution of telephone inquiries because problem resolution is an important predictor of customer satisfaction.7,8 Finally, we attempted to determine whether specific customer-service characteristics of reference laboratories were associated with greater resolution of telephone inquiries.
Definition of Terms
Primary Laboratory.—A laboratory that receives clinical specimens for testing directly after collection.
Reference Laboratory.—A laboratory that performs tests for and provides results to a primary laboratory.
Send-Out Test.—A test performed on a specimen dispatched to a reference laboratory from a primary laboratory.
Preanalytic Issues.—Questions about or problems with send-out tests that arise before test samples are assayed at a reference laboratory. These issues pertain to ordering, collecting, identifying, and transporting send-out specimens.
Postanalytic Issues.—Questions or problems (including delays and uncertainties) that arise after send-out test samples arrive at a reference laboratory. This definition includes both analytic issues that arise during testing and issues pertaining to results and reports.
Telephone Inquiry.—A single contact (eg, telephone call, fax transmission, or computer message) or attempted contact between the primary laboratory and the reference laboratory. One or more issues, patients, or tests may be the subject of an individual inquiry.
Institutions enrolled in the CAP Q-Probes Program participated in this study. The study was conducted and the data were handled in a manner similar to that previously described.9 Participants selected 1 reference laboratory to evaluate and provided the following information about their own practice setting on a preprinted input form: number of hospital beds, teaching status, monthly average number of tests their laboratory performs, monthly average number of tests their laboratory sends to reference laboratories, number of different reference laboratories their laboratory currently uses, name and type of reference laboratory (local hospital-affiliated laboratory, regional or university hospital-affiliated laboratory, state laboratory, federal laboratory, regional independent laboratory, or national independent laboratory) they chose to examine for this study, and number of tests their laboratory sent to that reference laboratory during the data collection period.
By making check marks on a detailed questionnaire, participants responded to specific inquiries concerning the following reference laboratory service characteristics (response options are listed in parentheses):
Distance between the participant's laboratory and the reference laboratory's processing site;
Method of entering clinical information into the reference laboratory's information system (written order form, computer, telephone, or no available entry method);
Method of specimen transport (mechanical delivery system, reference laboratory courier, sending laboratory courier, commercial express service, or mail);
Method of courier routing (direct, via collection depot for land transportation, via collection depot for air transportation, not known, or not applicable);
Pickup arrangements for routine specimens (scheduled, on demand, or both scheduled and on demand);
Usual response time for on-demand pickup of routine specimens (within 1 hour, 1–4 hours, within 8 hours, within 24 hours, within 48 hours, more than 48 hours, or not applicable);
Availability of transport for unscheduled STAT specimens (yes, no, or not applicable);
Usual response time for pickup of STAT specimens (within 1 hour, 1–4 hours, within 8 hours, within 24 hours, or not applicable);
Means of reporting results (by computer, teleprinter or fax, telephone, commercially expressed or courier-delivered print reports, or mail);
Availability of direct customer service telephone number (available toll free, available but not toll free, or not available);
Number of hours of telephone service on weekdays and on weekends (less than 8 hours, 8–16 hours, 16–24 hours, or not applicable);
Type of staff available by telephone on weekdays, evenings, nights, and weekends (clerks, medical technologists, pathologists, unknown, or not applicable).
For 2 months, or until 50 contacts had occurred, participants prospectively recorded the type and outcome of telephone inquiries from personnel in the participating laboratory to staff of the reference laboratory being monitored. Selecting from a prepared list of possible reasons (Table 1), participants recorded the reason for each telephone inquiry onto preprinted tally sheets. For each telephone inquiry, they also recorded whether the inquiry was promptly resolved by the first person answering the telephone; referred to someone else but resolved during the call; not resolved during the call but a return call was promised by the reference laboratory; not resolved during the call and the reference laboratory did not return the call as promised, necessitating a follow-up call by the primary laboratory; or not resolved and not referred. Telephone inquiries recorded as promptly resolved by the first person answering the telephone or referred to someone else but resolved during the call were categorized as resolved; all other inquiries were categorized as not resolved.
We compared the primary laboratories' demographics and practice characteristics and reference laboratories' customer service characteristics with the reference laboratories' abilities to resolve telephone inquiries. If a participant failed to answer a specific question about any of the monitored reference laboratory's service characteristics, that participant's data were excluded from the database for analysis of that question only. Wilcoxon and Kruskal-Wallis tests were used to assess differences between groups. We considered a P value of .05 or less to be statistically significant. A single statistician performed the data analysis. The identities of national reference laboratories were coded by the statistician to blind the authors assessing the results. This blinding remained in place throughout the study.
A total of 545 institutions participated in this study; 522 (95.8%) were located in 48 of the 50 states in the United States. The remaining 23 institutions (4.2%) were located in Canada (n = 14), Australia (n = 8), and New Zealand (n = 1). The institutions that provided information on number of hospital beds (529) were evenly divided into 3 groups, with 188 (35.5%) having 150 or fewer beds, 189 (35.7%) having 151 to 300 beds, and 152 (28.7%) having more than 300 beds. Two thirds of participants were classified as nonteaching institutions; the remaining third maintained teaching programs. The average number of tests that participants sent to reference laboratories ranged from 161 (10th percentile) to 1485 (90th percentile) per month, with the average primary laboratory sending 602 tests per month, which represented 1.5% of the median primary laboratory's total test volume. Participants reported using the services of 1 to 45 reference laboratories (median, 5 reference laboratories). The following types of reference laboratories were monitored: national reference laboratories (437 participants; 80.9%), regional reference laboratories (56; 10.4%), university-affiliated laboratories (24; 4.4%), local hospital laboratories (12; 2.2%), state government laboratories (8; 1.5%), and federal government laboratories (3; 0.6%). The distance between the participant and the processing site of the participant's reference laboratory was 10 miles or less for 102 participants (18.9%), 11 to 100 miles for 207 (38.4%), 101 to 500 miles for 126 (23.4%), and more than 500 miles for 104 (19.3%).
Table 2 shows the customer service characteristics of the reference laboratories monitored by the study participants. Most reference laboratories provided scheduled courier services (91%), STAT specimen pick-up services within 4 hours of receiving requests (82%), and toll-free numbers for customer service (91%). Most reference laboratories used clerks to respond to telephone inquiries (64%); reported results by teleprinter, fax, or computer link (92%); and had several types of backup reporting mechanisms. We were unable to identify significant differences in customer service characteristics between national and nonnational reference laboratories, nor could we identify significant differences in customer service characteristics among the 11 national reference laboratories that participants monitored for this study.
Participants tracked and classified 17 230 telephone contacts; the reasons for these contacts are listed in Table 1. There were 11714 inquiries (55%) concerning the preanalytic phase of testing and 9498 inquiries (45%) concerning the postanalytic phase; 1273 inquiries (6%) concerned both preanalytic and postanalytic issues. Preanalytically, the most frequent communication was a request for information about test selection (18%); postanalytically, the most frequent communication was a request for results (13%). Compared with customers of nonnational reference laboratories, customers of national reference laboratories more frequently inquired about test selection by 6% and about test results by 5%. The frequencies of inquiries concerning all other issues varied by no more than 2% between customers of national laboratories and those of nonnational laboratories. None of the differences observed between national and nonnational laboratories were statistically significant.
The outcomes of 14 017 telephone inquiries were characterized during this study. Overall, 11 031 telephone inquiries (78.7%) were resolved by personnel at the reference laboratories: 8452 (60.3%) were resolved by the first person answering the telephone and 2579 (18.4%) were resolved during the same call by someone else. Alternatively, 2986 inquiries (21.3%) were not resolved: for 2187 (15.6%) a return call was promised, 463 (3.3%) required a follow-up telephone call from the primary laboratory because the promised return call from the reference laboratory was never made, and 336 (2.4%) of calls were neither resolved nor referred.
For 533 primary laboratories that reported rates of inquiry resolution by the monitored reference laboratory, the rates varied considerably. The most successful 10th percentile of primary laboratories reported having 100% of their inquiries resolved. Other results were as follows: 75th percentile, 92.0% of calls resolved; 50th percentile, 81.0% of calls resolved; 25th percentile, 66.7% of calls resolved; and 10th percentile, 54.2% of calls resolved.
Primary laboratories that used the services of reference laboratories with 24-hour customer service had higher rates of inquiry resolution (median, 83.3%) than primary laboratories that used reference laboratories with less than 24-hour customer service (median, 77.7%; P = .0047). The rate of inquiry resolution did not significantly vary with any of the following 6 factors: number of hospital beds, primary laboratory in-house and send-out test volumes, distance between primary laboratory and reference laboratory, or reference laboratory customer service characteristics, including courier service, customer service personnel, and reporting mechanisms.
Table 3 shows, however, that the median rate of inquiry resolution did vary considerably among the 11 national reference laboratories monitored in this study, despite the similarities of service attributes listed in Table 2. The median resolution rate ranged from 55.0% to 90.2% (P < .0001).
We believe this to be the first study to compare levels of service among reference laboratories. The aspect of laboratory service that we chose to measure was the rate with which reference laboratory personnel were able to resolve questions telephoned to them by primary laboratory personnel. Overall, reference laboratory personnel resolved most telephone inquiries presented to them; however, the rate with which they resolved those telephone inquiries was not uniform among the reference laboratories. All inquiries were resolved for the 10th of primary laboratories reporting the highest rates of resolution; only slightly more than half of the inquiries were resolved for the 10th of primary laboratories reporting the lowest rates of resolution. We believe that resolution of at least 90% of inquiries, the level of performance we found by the top quarter of primary laboratories, is a reasonable performance benchmark for primary laboratories to set for reference laboratories.
In this study, only inquiries that were resolved during the first encounter with the reference laboratory were categorized as resolved. We believe this represents the level of reference laboratory service expected by most primary laboratory personnel. Included in the not-resolved group of inquiries were inquiries that resulted in a promised return telephone call by the reference laboratory. While some of these return telephone calls may have been made and may have subsequently resolved the problem, our study design did not allow us to capture this information. However, such eventual resolution entails, at least, an unwanted delay.
Of the reference laboratory customer service characteristics that we examined, only the presence of 24-hour telephone coverage was associated with a higher rate of telephone inquiry resolution. Because many laboratories operate 24 hours per day and inquiries are telephoned during all hours of operation, it is reasonable to conclude that reference laboratories that do not provide 24-hour telephone service solve fewer customer telephone inquiries than reference laboratories that can be reached at any time. However, it is unlikely that the hours of telephone coverage alone were responsible for the differences in inquiry resolution rate that we observed. For example, we found significant differences in resolution rate for the customers of the 11 national reference laboratories monitored, yet all 11 laboratories provided 24-hour telephone service.
We did not ask participants to record the reasons why telephone inquiries were not resolved. Others have suggested that factors influencing problem resolution include job knowledge, ability to take action, politeness and candor, speed of response, flexibility when presented with unusual requests, and commitment to promises.7,8,10 Objective studies are needed to determine whether these or other factors improve resolution of telephone inquiries.
Most primary laboratories participating in this study monitored the services of 1 of 11 national reference laboratories. Among these 11 laboratories we could not identify significant differences in customer service characteristics. On paper, all of these reference laboratories looked alike. However, the top-performing national reference laboratory resolved, on average, 90% of the telephone inquiries from its primary laboratory customers, whereas the bottom-performing laboratory resolved, on average, only slightly more than half of the telephone inquiries from its primary laboratory customers. These findings suggest that all reference laboratories do not provide the same level of customer service in telephone response. Given these findings, we believe that purchasers of reference laboratory services cannot take service standards for granted and should measure whether their reference laboratory meets objective performance standards like those for telephone inquiry resolution that we measured.
Comparing the rates with which laboratories resolve telephone inquiries must be done carefully. For example, a reference laboratory that extensively trains its employees and educates its customers may reduce the number of easily resolved telephone inquiries and inquiries overall, leaving a greater proportion of complex inquiries that are more difficult to resolve quickly. In contrast, a reference laboratory that provides less training and less education but receives more easily resolved telephone inquiries may have a higher resolution rate. Similarly, variation in the training and experience of primary laboratory personnel may affect problem resolution rates. After the Q-Probes study, we encouraged primary laboratories with lower rates of inquiry resolution to work with their reference laboratories to find ways first to reduce the necessity of telephone inquiries, then to increase the ability of reference laboratory personnel to resolve inquiries when calls must be made.
The reference laboratory service is a complex process; the quality of reference laboratory services is not characterized by a single outcome measure. Other measures of reference laboratory quality include the variety of tests available, ease of specimen transportation, frequency of lost specimens, characteristics of analytic performance, speed of turnaround time, readability of report format, competence of laboratory personnel, and availability of professional consultative services.11–15 Although the results of this study may stimulate primary laboratories to demand improved levels of service from their reference laboratories, this study of response to telephone inquiry can also serve as a template for evaluating other aspects of reference laboratory quality.
Statistical support was provided by Molly Walsh, PhD, College of American Pathologists, Northfield, Ill.
Reprints: Jane C. Dale, MD, Department of Laboratory Medicine and Pathology, Mayo Clinic, 200 First St SW, Rochester, MN 55905.