Abstract. Turnaround time for Papanicolaou (Pap) tests became an important service quality issue at our institution. We studied Pap test turnaround time using.
Anatomic Pathology / IMPROVING PAP TEST TURNAROUND TIME
Improving Pap Test Turnaround Time Using External Benchmark Data and Engineering Process Improvement Tools Thomas J. Persoon, MS,1,2 M. Sue Zaleski, MA,1 and Michael B. Cohen, MD1,3 Key Words: Papanicolaou test; Gynecologic smears; Process improvement; Turnaround time; Q-Probes; Action research; IDEF3
Abstract Turnaround time for Papanicolaou (Pap) tests became an important service quality issue at our institution. We studied Pap test turnaround time using engineering process improvement tools and benchmarked turnaround time against data published as a College of American Pathologists Q-Probes study. An IDEF3 process map revealed the complexity of the Pap test process and the opportunities for process improvement. We used these data and the actionresearch method to initiate changes in cytopathology laboratory operations with the goal of reducing turnaround time. Before intervention, mean Pap test turnaround time was highly variable; during a 6-month period, monthly means ranged from 2.5 to 10.8 days. A cycle time study conducted over a 2-week period validated these data. After system improvements were implemented, the monthly mean turnaround time decreased and became more consistent, with 11 of 12 months having a mean turnaround time of 3 days or less (range, 1.5-3.9 days). Our study illustrates the value of publishing Q-Probes data for use as external benchmarks and the benefits of using tools from other disciplines to improve laboratory processes.
© American Society for Clinical Pathology
The value of rapid turnaround times for Papanicolaou (Pap) tests has been the subject of debate. Yu and Gupta1 argued that rapid turnaround time for Pap tests is of doubtful value and that maintaining a steady-state turnaround time is adequate. On the other hand, Murray and Tantau2 noted that delay in any part of the health care process leads to patient dissatisfaction, and they argued that delay ought to be an unexpected rather than an expected annoyance. With respect to Pap tests, there are several customer satisfaction issues that are related to the value of prompt reporting: 1. Customers (patients) value the knowledge of a normal test result as providing reassurance of continuing health.3 2. Overall customer tolerance of delays in society is decreasing. Customers expect fast movement through the checkout lanes at the market, fast service at most restaurants, fast response from the Internet, and even fast service in changing the oil in their automobiles. Entire industries have arisen on the premise of providing faster service. 3. Women have expressed a preference for receiving reports of their Pap test results from their primary care provider, with an expected response time of 2 weeks or less.4 The popular press often sets expectations by publishing medical test turnaround times.5 Given these issues, it is not surprising that test turnaround time is an element in evaluations of the quality of health care services.6 A patient satisfaction survey conducted at our institution revealed that patients seen in the Department of Obstetrics and Gynecology were dissatisfied with the waiting time for their Pap test results. We therefore decided to measure and analyze the Pap test process at our institution. We sought to determine the root cause of the delays found and to implement process Am J Clin Pathol 2002;118:527-533
527
Persoon et al / IMPROVING PAP TEST TURNAROUND TIME
changes to improve this component of routine Pap test service quality. Because the process is complex, we applied tools and methods used by industrial engineers to analyze the process and suggest improvements.
Materials and Methods This study was conducted in an academic cytopathology laboratory with a combined gynecology and nongynecology volume of 48,000 slides per year. The staff consists of 4 cytotechnologists, 4 cytopathologists, and a clerk. A cytopreparation technician and a supervisor are allocated to cytopathology on a less than full-time basis. A key step in process improvement is the thorough and accurate documentation of the entire process. To document the process for Pap tests at our institution, we constructed an IDEF3 7 process schematic. Briefly, an IDEF3 process schematic is an object-oriented, information-rich method often used by industrial engineers to design and document complex processes. The process schematic is composed of a series of units of behavior (UOBs) that describe the activities performed on an object as it moves through a process. Each UOB documents the inputs, outputs, and resources required to perform the activity and any controls that may govern or limit the activity performance. A sample UOB is shown in ❚Figure 1❚. We interviewed staff members involved in the entire Pap test screening process, from sample collection to reporting of the results to the patient. The interviews were conducted by an industrial engineer (T.J.P.) who is familiar with process
Controls
C1
C2
Inputs
Outputs Stain slide
Labeled slide
UOB number
R1
Labeled and stained slide
R2
Resources
❚Figure 1❚ In an IDEF3 schematic, a unit of behavior (UOB) defines a specific activity, the inputs and outputs of the activity, the resources required to accomplish the activity, and the control measures that govern the activity. In the UOB for staining a slide, the input is the unstained, labeled slide, and the output is the labeled and stained slide. Resources required are R1 (laboratory technician) and R2 (slide staining machine). The controls for this UOB are C1 (laboratory procedure manual) and C2 (clock, since the staining process is a timed activity).
528
Am J Clin Pathol 2002;118:527-533
design. The information obtained from the interviews was used to create the IDEF3 process schematic. Although IDEF3 process schematics can be constructed using commercially available software or pen and paper, we chose to create the schematic of the Pap test process using various sized Post-it Notes (3M, St Paul, MN) and commercial white wrapping paper. This format enabled us to quickly reconfigure a part of the schematic when additional information became available and to have the schematic in a format that could be easily used to demonstrate the length and complexity of the process to groups of people. Information developed from the IDEF3 model was used to select segments of the Pap test process for further study. We were interested in obtaining detailed measurements of the cycle time (time needed for one object or set of objects to go through a process) for parts of the process governed by time controls. Because workers who know that their behaviors are being monitored often will change them to meet their perception of the behaviors desired by the monitoring agent, we selected monitoring tools that would be mostly invisible to workers. To measure the preanalytic process cycle time, the couriers who picked up Pap test slides from the Gynecology Clinic were given coded, self-adhesive labels to affix to the accompanying requisition at the time they picked up the slides. The couriers were chosen for this task because they made specimen pickups and deliveries on a wellcontrolled schedule and were considered to be only minor stakeholders in the overall test process, ie, they were unlikely to change their behavior while under observation. Once the requisitions and slides arrived in the cytopathology laboratory, the routine Pap test request was entered into the laboratory information system (LIS) (Cerner, Kansas City, MO) using the institution-assigned patient identification number. The LIS assigned a unique case accession number to the test request and automatically recorded the time accessioning occurred. After accessioning, the requisition was placed in a permanent file while the slide continued through the analytic process without further intervention. Later, we retrieved the requisitions from the files, decoded the labels to determine specimen pickup times, and mapped the pickup time to the accession number. These data then were entered into a computer spreadsheet. A total of 376 consecutive specimens from the gynecology clinics were obtained over 10 working days. After allowing several days for the Pap test process to be completed, we obtained the time of test completion (as measured by final result sign-out time recorded by the LIS) and entered it into the spreadsheet. Once test results have been reported, they are transmitted automatically by the LIS to the institution’s computerized medical record system and are available at any computerized medical record workstation. However, the IDEF3 schematic revealed that physicians used the printed “chart copy” report © American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE
Results The IDEF3 process schematic revealed that the Pap test process, from specimen collection to mailing of the result letter to the patient, had 74 UOBs. A portion of the * The Q-Probes study does not distinguish laboratories by size, location, or primary mission (ie, academic medical center vs community hospital). Our academic medical center laboratory fell between the 50th and 90th percentiles of the institutions participating in the Q-Probes study based on annual Pap test volume. Our laboratory did not participate in the Pap test Q-Probes study.
© American Society for Clinical Pathology
18 16 14
90% of samples Mean sample Theoretical end point
12 Days
of test results to generate a letter to patients informing them of a normal test result or to initiate a phone call to the patient if the test result was abnormal. We therefore used information from the IDEF3 schematic to determine the additional time required for the paper report to be printed and delivered to the physician. The reports were printed automatically at the same time every day by the LIS. They were delivered and distributed by the same couriers who pick up specimens, again on a well-controlled schedule. The time required for report printing and delivery was determined and entered into the spreadsheet for each case. Turnaround (cycle) time was calculated as the difference in calendar days (based on a 7 d/wk calendar) between specimen collection and report delivery. The mean and median turnaround times for the specimens in the detailed cycle time study were determined. We also instituted the practice of collecting turnaround time data available from the LIS. Summary data on turnaround time from accessioning to report filing were obtained from the LIS as a monthly mean. The total Pap test workload also was monitored to determine whether there were seasonal or other cyclic variations that could account for any changes observed. We used an organization development method called action research8 to effect change in the Pap test process. The IDEF3 process schematic, detailed cycle time data, and summarized monthly turnaround time data were presented to the cytopathology laboratory staff. The UOBs that could be responsible for substantial delays in cycle time were identified. In group sessions facilitated by the industrial engineer, the laboratory staff determined what factors affected the cycle (turnaround) time. The engineer led the staff to propose actions that could be taken by staff and by laboratory management personnel to improve the process and reduce or eliminate delays. During the period when the staff was working on identifying delays, the results of the College of American Pathologists Q-Probes study of gynecologic cytology turnaround time was published.9 The Q-Probes data were summarized in a graphic format by the engineer ❚Figure 2❚ and presented to the staff as external benchmark data.* The Q-Probes data and information from the customer satisfaction survey were used to set goals for turnaround time improvement.
10 8 6 4 2 0 Fastest 10% of Labs
Median Lab
Slowest 10% of Labs
❚Figure 2❚ The data from the College of American Pathologists Q-Probes study of routine Papanicolaou (Pap) test turnaround time, converted to graphic format. The fastest 10% of laboratories have very short median turnaround times and small differences between the median sample and the sample at the 90th percentile. These data indicate a tightly controlled process. In contrast, the slowest 10% of laboratories have not only a long median turnaround time but more variation in turnaround time, as demonstrated by the 8-day difference between the median and the 90th percentile sample. This may indicate a less tightly controlled process. Benchmark data should be considered a moving target. If all laboratories benchmark against these data and use them for continuous improvement, subsequent QProbes studies should reach a theoretical process improvement end point at which there is no differentiation between the fastest and the slowest laboratories and the turnaround time for all laboratories will approach zero. Data for this graph were extracted from Jones et al.9
schematic is shown in ❚Figure 3❚. Of the 74 UOBs, 49 were associated with the laboratory part of the process. The laboratory UOBs are listed in ❚Table 1❚. Twenty-four UOBs were determined to have a time-dependent control, with 16 of these associated with laboratory UOBs. A UOB with a time-dependent control was defined as one in which the activity occurred on a scheduled basis (ie, courier pickup and delivery times) or required a delay or waiting period (ie, slide drying after staining). The detailed cycle time study revealed that the fastest possible turnaround time for a Pap screening test collected during a 1:00 PM patient appointment was 48.5 hours (2.0 days) from specimen collection to delivery of the paper report to the physician’s mailbox (Table 1). The mean cycle (turnaround) time for the 376 specimens obtained during the detailed cycle time study was 9.4 calendar days (median, 9 calendar days; fastest, 7 days; slowest, 14 days), substantially longer than the fastest possible time and within the range of monthly turnaround times reported from the LIS ❚Figure 4❚. We were concerned Am J Clin Pathol 2002;118:527-533
529
Persoon et al / IMPROVING PAP TEST TURNAROUND TIME
C3 C2
C4
C2
Enter result
O1
R1 C1 C2 C3
Print copy
Affix electronic signature
31
32 R2
R1
R6
O1 R2
29
C2 C3 C7 O1
C2 O1
O2
45
O2
File slide
43 Normal
Normal View history
C6
Examine slide
O1
R7
X
O1
30
R8
X
Not normal
R1
R2
R2
Normal + history
R3 C2 C3 C5 O1
Place in hold basket
33 R2
Review history
O1
R5
O1
Examine slide
O1
35
34 R4
C2
R1
R3
R5
❚Figure 3❚ The IDEF3 (see text) process schematic for routine Papanicolaou (Pap) tests at our institution consisted of 74 units of behavior (UOBs). A portion of the IDEF3 process schematic is shown here. Objects (O), controls (C), and resources (R) are as follows: O1, prepared slide; O2, electronic result report; C1, cytotechnologist work schedule; C2, procedure manual; C3, clock; C4, preformatted laboratory information system (LIS) comment dictionary; C5, pathologists’ work schedule; C6, LIS print job schedule; C7, clerk work schedule; R1, computer workstation; R2, cytotechnologist; R3, microscope; R4, slide review holding box; R5, pathologist; R6, LIS printer; R7, filing clerk; R8, slide storage file. Clock and schedule controls are shown only for the UOBs in which a sequence of activities begins. An exclusive OR junction (X) is a point in a branched pathway at which an object may follow one and only one branch of the pathway. For a more complete explanation of the numbers in each group of boxes (eg, 29, 30), see Table 1.
with the mean turnaround time from collection to reporting and the relatively large range (7 days) of reporting times. Further attention was focused on the UOBs with timedependent controls. Inspection of the workstations where these UOBs were performed revealed a substantial number of Pap screening tests waiting for cytotechnologist screening. These were categorized as work in progress (WIP). At one time during the 10-day cycle time study period, the WIP was nearly 400 slides, representing approximately 3 days’ work for the entire cytotechnologist staff based on the average number of slides screened per day. A routine Pap test submitted at that time therefore would be delayed at least 3 days longer than if the WIP queue were zero. An additional substantial delay was discovered in the schedule for printing paper reports. The cytotechnologist staff was permitted to self-schedule, and most cytotechnologists chose to begin work early in the morning (6:00-7:00 AM) and leave in midafternoon (2:30-3:30 PM). However, the paper report for a Pap test specimen that was signed out at 7:30 AM would not be printed until 1:00 AM the following morning and would not be delivered to the 530
Am J Clin Pathol 2002;118:527-533
physician’s mailbox until 1:30 PM, adding another 1.3-day delay to result reporting. During the action-research phase of the study, the cytology laboratory staff developed an understanding of the necessity of improving Pap test turnaround time to meet customer expectations. The staff also came to understand the value of comparing their own results with benchmark data such as those available from the Q-Probes study. The laboratory’s physician customers expressed a desire for a Pap test turnaround time of fewer than 5 days. The cytology laboratory staff set an initial goal of achieving a mean turnaround time (receipt in laboratory to sign-out) of 4.4 days, the mean turnaround time of the median laboratory in the Q-Probes study. A longer term goal was to have a mean turnaround time equal to the fastest 10% of the laboratories in the QProbes study, 1.6 days. The staff determined 2 root causes of the large backlog of WIP. One root cause was the mix of responsibilities assigned to the cytotechnologist staff. In addition to Pap test screening, they were expected to microscopically examine nongynecologic specimens, perform quality assurance and quality control activities, participate in specimen preparation, © American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE
❚Table 1❚ Laboratory Units of Behavior (UOBs) for the Papanicolaou (Pap) Test* UOB No./Description 1/Collect sample from patient 2/Apply sample to slide and fix 3/Allow sample to dry 4/Transport sample to clinic pickup desk 5/Package slide for transport 6/Place slide in pickup basket 7/Transport slide to cytology laboratory 8/Place slide in cytology delivery basket 9/Review sample and apply acceptability rule 10/Check requisition for complete information 11/Label slide 12/Accession and transcribe relevant history 13/Verify accession number 14/Print work list 15/Place work list in holding basket 17/Load slide into staining rack 18/Place staining rack in ethanol 19/Place staining rack in autostainer 20/Place rack in xylene bath 21/Remove rack from xylene bath 22/Coverslip slide 23/Dry slide 24/Add color label 25/Verify labels with work list 26/Place slide on cytotechnologist’s tray 27/Place tray in holding basket 28/Cytotechnologist picks up 1 tray 29/View patient history in LIS 30/Examine slide 31/Enter preformatted comments 32/Affix electronic signature 33/Place slide in pathologist’s box 34/Review patient history in LIS (pathologist) 35/Examine slide (pathologist) 36/Circulate result sheet 37/Review slide and consult sheet 38/Add consult comments to sheet 40/Refer slide for second opinion
Time-Dependent Control
Clock Time
Day 1,1:00 PM 1:06 PM Drying time 1:12 PM 1:42 PM Staff work schedule 1:48 PM 1:48 PM Courier schedule 1:54 PM 2:12 PM Check-in clerk schedule 2:18 PM 2:24 PM 2:30 PM 2:36 PM 2:36 PM 2:37 PM 2:38 PM Laboratory assistant work schedule 3:00 PM Procedure-timed process 3:30 PM Procedure-timed process 4:12 PM 4:14 PM 4:15 PM 4:17 PM Procedure-timed process (overnight) 4:24 PM Day 2, 7:24 AM 7:24 AM Employee schedule 7:30 AM 7:30 AM Employee schedule 7:30 AM
Pathologists’ schedule Pathologists’ schedule
41/Examine slide (second opinion or quality assurance) 42/Affix electronic signature (second opinion)
Pathologists’ schedule
45/Print report “chart copy” 46/Sort reports 47/Place reports in hold basket 48/Transport reports to physician mailroom 49/Sort reports into physician mailboxes
LIS print schedule Employee schedule Courier schedule
Elapsed Time (h) 0 0.1 0.2 0.7 0.8 0.8 0.9 1.2 1.3 1.4 1.5 1.6 1.6 1.6 1.6 2.0 2.5 3.2 3.2 3.2 3.3 3.4 18.4 18.4 18.5 18.5 18.5
7:36 AM 7:42 AM 7:42 AM 7:48 AM 7:48 AM 9:00 AM 9:00 AM 9:12 AM 10:00 AM 10:06 AM 10:06 AM
18.6 18.7 18.7 18.8 18.8 20.0 20.0 20.2 21.0 21.1 21.1
11:00 AM
22.0
11:06 AM
22.1
1:00 PM Day 3, 7:30 AM 7:36 AM 11:00 AM 1:30 PM
36.0 42.5 42.6 46.0 48.5
Comment
This step involved a substantial amount of work in progress Normal slide pathway Normal slide pathway Nonnormal pathway Nonnormal pathway Nonnormal pathway Nonnormal pathway Nonnormal pathway Nonnormal pathway Slide with abnormal history in normal pathway Abnormal history or quality assurance Abnormal history or quality assurance
LIS, laboratory information system. * The IDEF3 model (see text) of the Pap test process before intervention and process modification contains 49 UOBs, with 47 UOBs falling on the critical path (essential for production of a Pap test result). UOBs having time controls are identified. The clock time and cumulative elapsed time is shown for a hypothetical Pap test sample collected at 1:00 PM that moves through the process with no unscheduled delays (assumes that each resource in the process is available when the sample reaches that UOB).
and assist in the maintenance of LIS databases. The staff determined that some tasks could be reassigned to noncytotechnologists, making additional time available for Pap test screening. The reassignment would occur as needed. As a result of the intervention, the cytotechnologists themselves began monitoring WIP on a daily basis and reassigning tasks when necessary. © American Society for Clinical Pathology
A second root cause of the large backlog of WIP was determined to be the result of an unexpected employee leave. While one cytotechnologist was on bereavement leave, inadequate provisions were made to reassign Pap test screening activities. The WIP created during the bereavement leave persisted for several weeks. Through the action-research process, the staff came to understand that, although specific Am J Clin Pathol 2002;118:527-533
531
Persoon et al / IMPROVING PAP TEST TURNAROUND TIME
12
Mean TAT, Days
10 8
Intervention period
6
Mean TAT for median lab = 4.4 days
4 2
Fastest 10% of labs report 90% of their results in 3 days
0
Mean TAT for fastest 10% of labs = 1.6 days
1/99
3/99
5/99
7/99
9/99
11/99 1/00 Month
3/00
5/00
7/00
9/00
11/00
❚Figure 4❚ Monthly mean Papanicolaou (Pap) test turnaround times (TATs) for the University of Iowa Hospitals and Clinics (UIHC) cytology laboratory are plotted on a graph that includes 3 benchmarks: (1) the mean TAT for the median laboratory in the College of American Pathologists (CAP) Q-Probes study,9 (2) the mean TAT for the laboratory at the fastest 10th percentile in the CAP Q-Probes study, and (3) the TAT benchmark for 90% of routine Pap tests for the laboratory at the fastest 10th percentile in the CAP Q-Probes study.
interruptions such as bereavement and unexpected medical leaves could not be predicted, some unexpected events would always occur. They developed a plan to reassign Pap test screening activities for expected staffing shortages such as vacations and for unexpected shortages. The plan included decreasing coverage at an off-site laboratory, directing nongynecologic specimens to the pathologists, and reducing the time for attending weekly educational conferences from 1 hour to 30 minutes. The plan was tested in the last quarter of 2000 when a staff member took parental leave. Pap test screening turnaround time remained constant during the leave period. Some laboratories experience seasonal variation in workload volume that, in turn, affects turnaround time. Our data ❚Figure 5❚ indicate that our turnaround time is independent of workload volume and seasonal effects.
The printing schedule for Pap test reports also was modified, so that reports were printed at 10:00 AM and distributed to the physicians’ mailboxes immediately after printing. This relatively minor change in operations also contributed to the improved turnaround time. Figure 4 shows the monthly mean Pap test turnaround time before, during, and after the action-research intervention. Before intervention, mean times ranged from 2.5 to 10.8 days. After intervention, mean times fell to between 1.5 and 3.9 days. Turnaround time has decreased steadily toward our target of being in the fastest 10% using the Q-Probes study criteria. Anecdotally, our physician customers report increased satisfaction with Pap test turnaround time.
Discussion 12
TAT, Days
10 8 6 4 2 0 1,800
1,900
2,000
2,100
2,200
2,300
2,400
2,500
No. of Tests
❚Figure 5❚ The relationship between Papanicolaou (Pap) test turnaround time (TAT) and workload volume before (circles) and after (squares) intervention. Pap test volume averaged 2,163 per month during the study period, with the maximum volume occurring in October 2000 (2,425) and the minimum occurring in December 2000 (1,812).
532
Am J Clin Pathol 2002;118:527-533
This study demonstrates the value of conducting and publishing studies of processes, such as the Q-Probes study of gynecologic cytology turnaround time. These studies serve as important benchmarks for laboratories committed to process improvement. In our opinion, these benchmarks represent moving targets, since comparison of performance is relative to a group statistic. As more laboratories strive to move themselves into the region of better performance, the group performance statistics change. Only the organizations committed to continuous process improvement can hope to maintain their performance ranking in such benchmark studies. Our study also demonstrates the value of using engineering tools such as the IDEF3 process schematics to analyze processes in health care laboratories. These tools are widely used in industrial and business environments but have not been © American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE
extensively used in health care. Before this study, the entire process for obtaining, transporting, analyzing, and reporting Pap tests in our institution was generally undocumented, and no single person could describe it in entirety. Both laboratory staff members and internal customers (clinic staff) were surprised to see the size and complexity of the process. It is our experience that health care professionals are trained to focus on minute details of their work. In doing so, they often fail to see or understand the organization and process systems in which they operate. The IDEF3 method serves to document the system, revealing system weaknesses such as process loops, hidden delays such as the report printing process described, and ineffective assignment of resources. Similarly, the action-research model of organization development, which is essentially the application of the scientific method to organization problems, also is underused in health care organizations but has great potential for solving problems. Application of this model requires a trained facilitator and willingness on the part of management to enlist employee participation and to support the conclusions obtained. In return, employees retain ownership of processes and assume responsibility for solving process problems. As part of process ownership, the cytotechnologists in our laboratory now monitor the WIP on a daily basis and take responsibility for scheduling and rescheduling themselves and support staff so that Pap test turnaround time goals can be met. Finally, this study also illustrates the substantial improvements in processes that can be made just by “picking the low-hanging fruit,” ie, making relatively easy changes in the system. The change in report printing times is an example of an easy change. As these easy changes are made, additional improvements become more difficult and require more sophisticated process-analysis tools and more detailed measurements of process properties.
© American Society for Clinical Pathology
From the Departments of 1Pathology, 2Industrial Engineering, and 3Epidemiology, University of Iowa, Iowa City. Address reprint requests to Dr Cohen: Dept of Pathology, University of Iowa, 200 Hawkins Dr, Iowa City, IA 52242. Acknowledgments: We acknowledge the assistance of the Pathology Department courier staff and the cytology laboratory staff in collecting data, and Joel Carl, MA, for help in preparing the figures.
References 1. Yu GH, Gupta PK. The pathologic obsession with turnaround time in gynecologic cytology: is it warranted [editorial]? Diagn Cytopathol. 1998;19:321-322. 2. Murray M, Tantau C. Must patients wait [editorial]? Jt Comm J Qual Improv. 1998;24:423-425. 3. Schofield MJ, Sanson-Fisher R, Halpin S, et al. Notification and follow-up of Pap test results: current practice and women’s preferences. Prev Med. 1994;23:276-283. 4. Andreasen LJ, Holund B, Jeune B, et al. Screening against cervical cancer: experiences, attitudes and knowledge of women in the county of Funen. Ugeskr Laeger. 1998;160:405409. 5. Test turnaround time. Ladies Home J. September 1999:80. 6. Haddad S, Potvin L, Roberge D, et al. Patient perception of quality following a visit to a doctor in a primary care unit. Fam Pract. 2000;17:21-29. 7. Mayer RJ, Menzel CP, Painter MK, et al. Information Integration for Concurrent Engineering (IICE): IDEF3 Process Description Capture Method Report. College Station, TX: Knowledge Based Systems; 1995. 8. French W, Bell C Jr. Organization Development: Behavioral Science Interventions for Organization Improvement. Englewood Cliffs, NJ: Prentice-Hall; 1990. 9. Jones BA, Valenstein PN, Steindel SJ. Gynecologic cytology turnaround time: a College of American Pathologists QProbes study of 371 laboratories. Arch Pathol Lab Med. 1999;123:682-686.
Am J Clin Pathol 2002;118:527-533
533
