Quality Assurance in Health Can, Vol. 4, No. 3, pp. 245-256, 1992 Printed in Great Britain
1040-6166/92 S5.00 + 0.00 Pergaroon Press Ltd
A NATIONWIDE QUALITY ASSURANCE PROGRAM CAN DESCRIBE STANDARDS FOR THE PRACTICE OF PATHOLOGY AND LABORATORY MEDICINE
Clinical Laboratories UCLA, Los Angeles, CA, USA •Department of Pathology Lake Forest Hospital Lake Forest, IL, USA tDepartment of Pathology Tucson VA Medical Center Tucson, AZ, USA t Department of Pathology Henry Ford Hospital Detroit, MI, USA §Prism Associates, Inc. Atlanta, GA, USA ||Royal Jubilee Hospital Dept. of Laboratory Medicine 1900 Fort Street Victoria, BC V8R 1J8, Canada (First submitted 7 November 1991, accepted after revision 16 January 1992)
An important component of quality assessment is the analysis of peer group comparisons, although little data are available for evaluation. We developed and tested six interinstitutional quality indicators related to Pathology and Laboratory Medicine among 36 institutions. Results showed that the mean frequency of intraoperative frozen section consultations (6.0%), sensitivity of fine needle aspiration cytology diagnosis (87%), nosocomial infections (5.0%) and average cross-match to transfusion ratio (2.1 %) was comparable with previous studies, but the range of values was large. The median stat laboratory turnaround time of approximately 1 hrfor CSF cell count, glucose, protein and gram smear was considerably longer than expected from previous investigations, and was longer for larger institutions. Analysis of serious laboratory reporting errors showed the lowest number detected by individuals working in transfusion medicine, and highest numbers among hematology workers. We conclude that interinstitutional comparison of data from quality assurance programs can be used to describe performance standards related to the quality and effectiveness of care. Key words: Quality assurance, quality improvement, quality frozen sections, nosocomial infections, test turnaround times, report errors, blood usage, fine needle aspiration cytology. 245
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/245/1797170 by guest on 25 October 2019
Peter J. Howanitz, Gerald G. Hoffman,* Ron B. Schifman.t Richard J. Zarbo,* Steven J. Steindel§ and Keith Walker||
246
P. J. Howanitz et al.
INTRODUCTION
MATERIALS AND METHODS Members of the College of American Pathologists (CAP) Quality Assurance Service (QAS) Committee developed and tested six Quality Assurance modules in hospitals in which they worked. Modules in Anatomic Pathology and Laboratory Medicine were combined and represented a major part of a comprehensive Quality Assurance program, Q-Probes®, offered for subscription by the CAP beginning in January 1989. Active members of the CAP and others were solicited for participation of their institutions in a "pilot study" of these six modules prior to general enrolment of hospital Pathology and Laboratory Medicine Departments in Q-Probes. Of the 56 institutions whose pathologist agreed to participate, 36 returned data describing performance. Participant data were evaluated, and data were not included for study when input forms were incompletely or incorrectly filled out.
Reprint requests should be addressed to: Peter J. Howanitz, M.D., Clinical Laboratories, UCLA Medical Center, 10833 LeConte Avenue, Los Angeles, CA 90024, USA
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/245/1797170 by guest on 25 October 2019
Quality Assurance is a surveillance process designed to guarantee that all services involved in the delivery of patient care have been accomplished in a manner appropriate to maintain excellence. Since 1951, the Joint Commission for Accreditation of Health Care Organizations (JCAHO) has worked to improve quality in health care settings. Their most recent emphasis called "The Agenda for Change" has been aimed at determining the relationship between patient outcomes and the structures and processes which produce these results [1]. According to a 10-step monitoring and evaluation process, appropriate quality indicators of patient care are developed, and investigations occur when a standard or "threshold for action" is exceeded [2]. Although in the future, the JCAHO may recommend a series of quality indicators for Pathology and Laboratory Medicine, currently each institution in the US identifies thresholds for action, evaluates results and takes corrective action independent of any comparative norm. In the past, Quality Assurance procedures in Laboratory Medicine primarily involved inclusion of control samples that mimic patient specimens in each analytical run [3]. This practice, which related only to the analytical determination, has been documented by interlaboratory comparisons of data from daily quality control and proficiency programs, but unfortunately has neglected pre-analytical and postanalytical components of the test ordering process. It also has no analogous counterpart in Anatomic Pathology. These daily quality control and proficiency programs measure the quality of performance of individual participant laboratories, measure the state-of-the-art of clinical practice and serve as an educational stimulus for laboratory improvement. We describe a modular Quality Assurance program of the College of American Pathologists, Q-Probes, which identifies key quality indicators in Pathology and Laboratory Medicine, tests six modules in volunteer facilities and determines the state-of-the-art performance using comparison of results similar to laboratory quality control and proficiency testing programs.
Nationwide QA Describes Standards
247
Frozen Section Correlation We asked participants to collect data prospectively for 4 weeks on the number of surgical pathology accessions and the number of intraoperative surgical pathology frozen section consultations. Institutions were divided by size; large institutions were defined as performing more than 1000 pathology accessions during the study period whereas small institutions were defined as performing less than 1000 accessions. Fine Needle Aspiration Cytohistoligical Correlation For fine needle aspiration of the lung, diagnoses made during 6 months from cytological tissue were compared retrospectively with the final tissue diagnoses. In cases of discordance between the two diagnoses, the tissue from the fine needle aspiration and thefixedsections were reinterpreted. If discordance still occurred, the medical record and the patient's clinical course were reviewed to obtain the correct tissue diagnosis for FNAC performance. The initial FNAC diagnoses were compared with the correct tissue diagnosis. Interpretations from each case were collated and overall sensitivity and specificity determined. Cerebrospinal Fluid (CSF) Turnaround Time For turnaround time, participants in pilot laboratories recorded the times when each CSF specimen was logged into the laboratory and when each result was reported. Turnaround time was defined as the time difference between specimen logging and reporting of results. Laboratories were required to collect data from consecutive CSF specimens until they reached 30 or until the 4-week study ended, whichever occurred first. Nosocomial Infection Rates For nosocomial infection rates, participants responded to a questionnaire about their guidelines for frequency and extent of surveillance, the mechanism of case finding and how nosocomial infection data is expressed in their institutions. Data on nosocomial infection rates at each institution were collected retrospectively for one year, when totaled by rate for teaching and non-teaching institutions.
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/245/1797170 by guest on 25 October 2019
The Anatomic Pathology modules evaluated were: (1) intraoperative frozen section consultation and (2) concordance between the diagnosis obtained from fine needle aspiration cytology (FNAC) of the lung and the final histologic diagnosis obtained from fixed tissue. The Laboratory Medicine modules included: (3) turnaround time of stat cerebrospinal fluid glucose, protein, cell count and gram stain results: (4) nosocomial infection rates; (5) quantitation and impact on patient care of corrected laboratory reports; and (6) red blood cell utilization. Each module required collection of institutional demographic material as well, including bed size, type of facility and number of pathologists involved in various activities. For each module, data from each institution was averaged, then the average for each institution was used to calculate mean rates for percentile rank.
248
P. J. Howanitz et al. TABLE 1.
Interinstttntioiial frozen section comparison
Percentile rank
0.0 5.0
Rate (%)
0.3 0.8 2.1 3.2 5.6 8.3 10.4 11.7 19.2
Institutions are ranked by percentile dependent on frozen section rates.
Laboratory Error Rates The changes made during a month to laboratory reports after those reports had been finalized and made available for patient care were quantitated. Changes were classified based on the impact on patient care: Category A errors were those where the physician had evaluated a result and treatment was based on that result; Category B errors were those which were unlikely to affect patient care or were potential Category A errors which were detected before the physician's review; whereas, Category C were minor clerical errors and cosmetic corrections of the report. Examples of the three types of errors were listed for participants and included an A error [a plasma glucose value of 15 mg/dL (0.8 mmol/L) instead of 150 mg/dL (8.0 mmol/L) causing the patient to receive an intravenous infusion of glucose], a B error (the same erroneous glucose result discovered before the patient was treated) and a C error (changing the cell type on a gram stain report from "polyps" to "white blood cells"). The number of changes in the laboratory report were expressed in terms of the number of full time equivalents (FTEs) employed 40 hr per week in each laboratory discipline: hematology, chemistry, microbiology and transfusion medicine (blood banking). Participants were required to include immunology as part of microbiology, urinalysis and toxicology as part of chemistry and coagulation as part of hematology. Red Blood Cell Utilization The units of red blood cells (RBCs), frozen RBCs and leukocyte-poor RBCs requested by physicians were totaled for each laboratory and divided by the total number transfused for that laboratory during the previous year. RESULTS Table 1 shows the frequency of intraoperative frozen section consultations as a percentage total surgical pathology procedures in 34 participating Surgical Pathology
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/245/1797170 by guest on 25 October 2019
10.0 25.0 50.0 75.0 90.0 95.0 100.0
Nationwide QA Describes Standards TABLE 2.
249 Diagnostic performance of fine needle aspiration 579 232 11 87% 98% 76% 99% 70% 90%
Cases Histological correlations Deferred diagnoses Mean sensitivity Mean specificity Prevalence of malignancy Predictive value of positive Predictive value of negative Efficiency
TABLE 3. 250Beds
1040-6166/92 S5.00 + 0.00 Pergaroon Press Ltd
A NATIONWIDE QUALITY ASSURANCE PROGRAM CAN DESCRIBE STANDARDS FOR THE PRACTICE OF PATHOLOGY AND LABORATORY MEDICINE
Clinical Laboratories UCLA, Los Angeles, CA, USA •Department of Pathology Lake Forest Hospital Lake Forest, IL, USA tDepartment of Pathology Tucson VA Medical Center Tucson, AZ, USA t Department of Pathology Henry Ford Hospital Detroit, MI, USA §Prism Associates, Inc. Atlanta, GA, USA ||Royal Jubilee Hospital Dept. of Laboratory Medicine 1900 Fort Street Victoria, BC V8R 1J8, Canada (First submitted 7 November 1991, accepted after revision 16 January 1992)
An important component of quality assessment is the analysis of peer group comparisons, although little data are available for evaluation. We developed and tested six interinstitutional quality indicators related to Pathology and Laboratory Medicine among 36 institutions. Results showed that the mean frequency of intraoperative frozen section consultations (6.0%), sensitivity of fine needle aspiration cytology diagnosis (87%), nosocomial infections (5.0%) and average cross-match to transfusion ratio (2.1 %) was comparable with previous studies, but the range of values was large. The median stat laboratory turnaround time of approximately 1 hrfor CSF cell count, glucose, protein and gram smear was considerably longer than expected from previous investigations, and was longer for larger institutions. Analysis of serious laboratory reporting errors showed the lowest number detected by individuals working in transfusion medicine, and highest numbers among hematology workers. We conclude that interinstitutional comparison of data from quality assurance programs can be used to describe performance standards related to the quality and effectiveness of care. Key words: Quality assurance, quality improvement, quality frozen sections, nosocomial infections, test turnaround times, report errors, blood usage, fine needle aspiration cytology. 245
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/245/1797170 by guest on 25 October 2019
Peter J. Howanitz, Gerald G. Hoffman,* Ron B. Schifman.t Richard J. Zarbo,* Steven J. Steindel§ and Keith Walker||
246
P. J. Howanitz et al.
INTRODUCTION
MATERIALS AND METHODS Members of the College of American Pathologists (CAP) Quality Assurance Service (QAS) Committee developed and tested six Quality Assurance modules in hospitals in which they worked. Modules in Anatomic Pathology and Laboratory Medicine were combined and represented a major part of a comprehensive Quality Assurance program, Q-Probes®, offered for subscription by the CAP beginning in January 1989. Active members of the CAP and others were solicited for participation of their institutions in a "pilot study" of these six modules prior to general enrolment of hospital Pathology and Laboratory Medicine Departments in Q-Probes. Of the 56 institutions whose pathologist agreed to participate, 36 returned data describing performance. Participant data were evaluated, and data were not included for study when input forms were incompletely or incorrectly filled out.
Reprint requests should be addressed to: Peter J. Howanitz, M.D., Clinical Laboratories, UCLA Medical Center, 10833 LeConte Avenue, Los Angeles, CA 90024, USA
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/245/1797170 by guest on 25 October 2019
Quality Assurance is a surveillance process designed to guarantee that all services involved in the delivery of patient care have been accomplished in a manner appropriate to maintain excellence. Since 1951, the Joint Commission for Accreditation of Health Care Organizations (JCAHO) has worked to improve quality in health care settings. Their most recent emphasis called "The Agenda for Change" has been aimed at determining the relationship between patient outcomes and the structures and processes which produce these results [1]. According to a 10-step monitoring and evaluation process, appropriate quality indicators of patient care are developed, and investigations occur when a standard or "threshold for action" is exceeded [2]. Although in the future, the JCAHO may recommend a series of quality indicators for Pathology and Laboratory Medicine, currently each institution in the US identifies thresholds for action, evaluates results and takes corrective action independent of any comparative norm. In the past, Quality Assurance procedures in Laboratory Medicine primarily involved inclusion of control samples that mimic patient specimens in each analytical run [3]. This practice, which related only to the analytical determination, has been documented by interlaboratory comparisons of data from daily quality control and proficiency programs, but unfortunately has neglected pre-analytical and postanalytical components of the test ordering process. It also has no analogous counterpart in Anatomic Pathology. These daily quality control and proficiency programs measure the quality of performance of individual participant laboratories, measure the state-of-the-art of clinical practice and serve as an educational stimulus for laboratory improvement. We describe a modular Quality Assurance program of the College of American Pathologists, Q-Probes, which identifies key quality indicators in Pathology and Laboratory Medicine, tests six modules in volunteer facilities and determines the state-of-the-art performance using comparison of results similar to laboratory quality control and proficiency testing programs.
Nationwide QA Describes Standards
247
Frozen Section Correlation We asked participants to collect data prospectively for 4 weeks on the number of surgical pathology accessions and the number of intraoperative surgical pathology frozen section consultations. Institutions were divided by size; large institutions were defined as performing more than 1000 pathology accessions during the study period whereas small institutions were defined as performing less than 1000 accessions. Fine Needle Aspiration Cytohistoligical Correlation For fine needle aspiration of the lung, diagnoses made during 6 months from cytological tissue were compared retrospectively with the final tissue diagnoses. In cases of discordance between the two diagnoses, the tissue from the fine needle aspiration and thefixedsections were reinterpreted. If discordance still occurred, the medical record and the patient's clinical course were reviewed to obtain the correct tissue diagnosis for FNAC performance. The initial FNAC diagnoses were compared with the correct tissue diagnosis. Interpretations from each case were collated and overall sensitivity and specificity determined. Cerebrospinal Fluid (CSF) Turnaround Time For turnaround time, participants in pilot laboratories recorded the times when each CSF specimen was logged into the laboratory and when each result was reported. Turnaround time was defined as the time difference between specimen logging and reporting of results. Laboratories were required to collect data from consecutive CSF specimens until they reached 30 or until the 4-week study ended, whichever occurred first. Nosocomial Infection Rates For nosocomial infection rates, participants responded to a questionnaire about their guidelines for frequency and extent of surveillance, the mechanism of case finding and how nosocomial infection data is expressed in their institutions. Data on nosocomial infection rates at each institution were collected retrospectively for one year, when totaled by rate for teaching and non-teaching institutions.
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/245/1797170 by guest on 25 October 2019
The Anatomic Pathology modules evaluated were: (1) intraoperative frozen section consultation and (2) concordance between the diagnosis obtained from fine needle aspiration cytology (FNAC) of the lung and the final histologic diagnosis obtained from fixed tissue. The Laboratory Medicine modules included: (3) turnaround time of stat cerebrospinal fluid glucose, protein, cell count and gram stain results: (4) nosocomial infection rates; (5) quantitation and impact on patient care of corrected laboratory reports; and (6) red blood cell utilization. Each module required collection of institutional demographic material as well, including bed size, type of facility and number of pathologists involved in various activities. For each module, data from each institution was averaged, then the average for each institution was used to calculate mean rates for percentile rank.
248
P. J. Howanitz et al. TABLE 1.
Interinstttntioiial frozen section comparison
Percentile rank
0.0 5.0
Rate (%)
0.3 0.8 2.1 3.2 5.6 8.3 10.4 11.7 19.2
Institutions are ranked by percentile dependent on frozen section rates.
Laboratory Error Rates The changes made during a month to laboratory reports after those reports had been finalized and made available for patient care were quantitated. Changes were classified based on the impact on patient care: Category A errors were those where the physician had evaluated a result and treatment was based on that result; Category B errors were those which were unlikely to affect patient care or were potential Category A errors which were detected before the physician's review; whereas, Category C were minor clerical errors and cosmetic corrections of the report. Examples of the three types of errors were listed for participants and included an A error [a plasma glucose value of 15 mg/dL (0.8 mmol/L) instead of 150 mg/dL (8.0 mmol/L) causing the patient to receive an intravenous infusion of glucose], a B error (the same erroneous glucose result discovered before the patient was treated) and a C error (changing the cell type on a gram stain report from "polyps" to "white blood cells"). The number of changes in the laboratory report were expressed in terms of the number of full time equivalents (FTEs) employed 40 hr per week in each laboratory discipline: hematology, chemistry, microbiology and transfusion medicine (blood banking). Participants were required to include immunology as part of microbiology, urinalysis and toxicology as part of chemistry and coagulation as part of hematology. Red Blood Cell Utilization The units of red blood cells (RBCs), frozen RBCs and leukocyte-poor RBCs requested by physicians were totaled for each laboratory and divided by the total number transfused for that laboratory during the previous year. RESULTS Table 1 shows the frequency of intraoperative frozen section consultations as a percentage total surgical pathology procedures in 34 participating Surgical Pathology
Downloaded from https://academic.oup.com/intqhc/article-abstract/4/3/245/1797170 by guest on 25 October 2019
10.0 25.0 50.0 75.0 90.0 95.0 100.0
Nationwide QA Describes Standards TABLE 2.
249 Diagnostic performance of fine needle aspiration 579 232 11 87% 98% 76% 99% 70% 90%
Cases Histological correlations Deferred diagnoses Mean sensitivity Mean specificity Prevalence of malignancy Predictive value of positive Predictive value of negative Efficiency
TABLE 3. 250Beds
