From customer satisfaction survey to corrective actions in laboratory services in a university hospital

International Journal for Quality in Health Care 2006; Volume 18, Number 6: pp. 422 428 Advance Access Publication: 26 September 2006 From customer satisfaction survey to corrective actions in laboratory services in a university hospital 10.1093/intqhc/mzl050 PAULA I. OJA 1,2, TIMO T. KOURI 1,2 AND ARTO J. PAKARINEN 1,2 1 Department of Clinical Chemistry, University of Oulu and 2 Laboratory, Oulu University Hospital, Oulu, Finland Abstract Objective. To find out the satisfaction of clinical units with laboratory services in a university hospital, to point out the most important problems and defects in services, to carry out corrective actions, and thereafter to identify the possible changes in satisfaction. Setting and study participants. Senior physicians and nurses-in-charge of the clinical units at Oulu University Hospital, Finland. Design. Customer satisfaction survey using a questionnaire was carried out in 2001, indicating the essential aspects of laboratory services. Customer-specific problems were clarified, corrective actions were performed, and the survey was repeated in 2004. Results. In 2001, the highest dissatisfaction rates were recorded for computerized test requesting and reporting, turnaround times of tests, and the schedule of phlebotomy rounds. The old laboratory information system was not amenable to major improvements, and it was renewed in 2004 05. Several clinical units perceived turnaround times to be long, because the tests were ordered as routine despite emergency needs. Instructions about stat requesting were given to these units. However, no changes were evident in the satisfaction level in the 2004 survey. Following negotiations with the clinics, phlebotomy rounds were re-scheduled. This resulted in a distinct increase in satisfaction in 2004. Conclusions. Satisfaction survey is a screening tool that identifies topics of dissatisfaction. Without further clarifications, it is not possible to find out the specific problems of customers and to undertake targeted corrective actions. Customer-specific corrections are rarely seen as improvements in overall satisfaction rates. Keywords: customer satisfaction, laboratory services, quality management, satisfaction survey Customer orientation has got increasing attention in health care. The implementation of quality standards, such as ISO 15189 [1] and ISO 17025 [2], and the use of management systems, e.g. Balanced Scorecard [3], in clinical laboratories have further emphasized the customer perspective in the improvement of laboratory service. It is a challenge for the clinical laboratory management to reach and utilize the customer perspective. A generally used method to obtain customer feedback is to conduct a satisfaction survey. Satisfaction surveys provide satisfaction ratings. However, the underlying problems behind dissatisfaction may not be revealed by surveys alone. Additionally, clarifications with selected customers are needed. The purpose of this study was to assess the satisfaction of clinical units of Oulu University Hospital, to find out the problems causing dissatisfaction, and to correct defects in the laboratory process. Methods The satisfaction survey was carried out in the clinics of Oulu University Hospital, Oulu, Finland. The hospital, with nearly 1000 beds and about 1000 outpatient visits/weekday, includes all main clinical specialties. The clinical chemistry laboratory provides services in clinical chemistry, hematology, and nuclear medicine with a total of 2.5 million investigations annually, 70% of the requests coming from the hospital. Phlebotomy and analytical services are provided by the laboratory 24 hours a day. Practically, all venous and skin puncture blood specimens, both scheduled and non-scheduled, are taken by medical laboratory technologists. The laboratory has ISO 17025 accreditation for most routine tests. The satisfaction survey was carried out using a questionnaire planned by the chief physician, the associate chief physician, both specialists with long experience in clinical Address reprint requests to Paula I. Oja, Laboratory, Oulu University Hospital, PO Box 500, 90029 OYS, Finland. E-mail: paula.oja@ppshp.fi International Journal for Quality in Health Care vol. 18 no. 6 The Author 2006. Published by Oxford University Press on behalf of International Society for Quality in Health Care; all rights reserved 422

Customer satisfaction of laboratory services chemistry, and the planning officer of the laboratory [4,5]. The statements were carefully designed to cover the essential aspects of laboratory services at Oulu University Hospital (Table 1) [6]. Special attention was paid to the statements wording and the number of ordinal scale categories to obtain valid information on subjective questions [7]. In addition, the design of the questionnaire was reviewed by the managing board of the laboratory. The respondent letter was signed by the chief physician [8]. The respondents were asked to rate their satisfaction on a 5-point Likert scale: 1 (strongly agree), 2 (agree), 3 (neither agree nor disagree), 4 (disagree), and 5 (strongly disagree) [5]. The respondents were also instructed to use not applicable if appropriate. In addition to these closed statements, the respondents were asked an open-ended question: What is the most important problem in laboratory services?. Questionnaires on paper were sent to the senior physicians and nurses-in-charge of the in-patient and outpatient units of the clinics using internal mail of the hospital. They were also asked to take notice of the opinions of their medical and nursing personnel. In addition, the respondents were asked to give their contact information, to make them accountable for their responses, and to make it possible to identify the units specific problems [5]. Identical questionnaires were sent out in 2001 and 2004. In 2001, 136 questionnaires were sent to 68 clinical units, and in 2004, 144 questionnaires were sent to 72 units. Percentage distributions of the responses were used to present the data. As a screening tool, a cumulative percentage of the two disagreement levels (disagree and strongly disagree) of 20% was considered to be a high level of dissatisfaction. These items were selected for further problem clarification and possible corrective actions. The significance between the differences in the classified satisfaction rates in 2001 and 2004 was tested by using the chi-squared test. The responses to the open-ended question were analyzed by content analysis with calculated frequencies by two coders [9]. On the basis of the 2001 questionnaire, 15 clinical units were visited in 2002 to clarify their specific problems and to negotiate possible corrective actions needed. Written records of the negotiations were produced. Their contents were analyzed twice by one coder [9]. Corrective actions were performed on the basis of the information obtained from the questionnaire and from the negotiations with the units. Results Response rates In 2001, 70 respondents returned questionnaires from 54 clinical units. In 2004, 91 respondents returned questionnaires from 64 clinical units. The response rate of the clinical units to the closed statements was 79.4% in 2001 and 88.9% in 2004. The responding units requested 87.7 and 95.3% of the total number of tests performed by our Laboratory for the hospital units in 2001 and 2004, respectively. Survey in 2001 The distributions of the responses to the statements on laboratory services are summarized in Table 1. In 2001, the highest percentage of disagreement (35.9%) was related to missing test results. In addition, there was a high percentage of dissatisfaction with the laboratory information system when reviewing the laboratory results (30.8%), and with the turnaround times of both stat tests (28.4%) and routine tests for inpatients (26.5%). The respondents needed additional instructions on the preparation of patients for laboratory tests (27.3%) and on the collection and handling of samples (27.8%). Approximately 25.9% needed additional consultations by laboratory physicians. The use of Laboratory Users Handbook was criticized by 22.4% of the respondents, and about 21% were not satisfied with the schedule of phlebotomy rounds. We compared the responses of the senior physicians and the nurses-in-charge from the units that gave separate answers. The only significant difference in dissatisfaction concerned the use of laboratory information system when requesting laboratory tests and reviewing laboratory results. The most important problems in laboratory services, based on the responses to the open-ended question in 2001, are summarized in Table 2. The respondents reported a total of 103 problems (range 0 6, average 1.8 problems/response). The problems most frequently reported pertained to computerized test requests and reporting, phlebotomy services, and turnaround times of tests. Many respondents perceived the electronic information supporting the ordering, reporting, and interpretation of laboratory tests as being insufficient and electronic requesting as too complicated. A variety of problems were related to the in-patient phlebotomy services. There were not enough unscheduled phlebotomy services and scheduled rounds. Some pediatric units pointed out the shortage of out-of-hours phlebotomy services. The criticism of turnaround times was mainly directed at the regular daytime services concerning single tests for certain clinics, e.g. C-reactive protein, liver function tests and creatinine clearance for the outpatient unit of oncology, C-reactive protein for the outpatient unit of surgery, urine amylase for an inpatient unit of surgery, prostate-specific antigen and bone scintigraphy for the outpatient unit of urology, and leukocyte differential count for in-patient units of pediatrics. Corrective actions Missing test results. At the negotiations with the selected clinics, the respondents were asked to explain what they meant by a missing test result (Table 1, statement 10a). Most often the results were considered missing if they were received later than expected for the intended purpose. Some clinical units did not use stat requests in urgent cases but expected a delay of urgent tests in all cases. They were informed about the turnaround times of laboratory tests, and they were asked to use stat requests in urgent cases. A few clinical units expected that the routine morning test results for in-patients could already be available for the early morning medical rounds. 423

P. I. Oja et al. Table 1 Distribution of responses to the statements exploring clinics view of laboratory services in 2001 and 2004 Statements assessed Year Number of answers Strongly agree/ agree (%) Neither agree nor Disagree/ strongly P-value 1... 1. Service attitude of the laboratory personnel is good 2001 68 92.7 4.4 2.9 2004 87 88.5 5.7 5.7 NA 2. Laboratory informs adequately of the changes in services 2001 68 85.2 8.8 8.9 2004 91 90.2 5.5 4.4 NA 3a. Laboratory Users Handbook is fit for use 2001 67 61.2 16.4 22.4 2004 90 64.4 14.4 21.1 0.91 3b. Updated test record is fit for use 2001 61 78.7 11.5 9.8 2004 86 72.1 9.3 18.6 0.33 3c. Reference value booklet is fit for use 2001 49 61.2 32.7 6.1 2004 79 64.5 27.8 7.6 0.83 4. We are satisfied with the laboratory information system a. when requesting laboratory tests 2001 58 63.8 17.2 19.0 2004 76 50.0 13.2 36.8 0.08 b. when reviewing laboratory results in electrical patient records 2001 68 55.9 13.2 30.8 2004 88 71.6 10.2 18.2 0.11 c. when receiving cumulative reports 2001 64 67.2 12.5 20.3 2004 85 71.7 12.9 15.3 0.73 5. Requisition and delivery of blood components are fluent 2001 49 71.5 12.2 16.3 2004 62 74.1 11.3 14.5 0.95 6. The price of laboratory tests does not limit requests a. in clinical chemistry and hematology 2001 52 76.9 11.5 11.5 2004 73 78.1 13.7 8.2 0.79 b. in nuclear medicine 2001 43 76.7 9.3 13.9 2004 62 66.2 17.7 16.1 0.42 7. We are satisfied with the schedule of phlebotomy rounds a. in the daytime 2001 38 73.7 5.3 21.1 2004 60 88.3 10.0 1.7 0.005 b. during the emergency hours 2001 35 68.5 11.4 20.0 2004 58 82.7 8.6 8.6 0.23 8. We do not need additional instructions a. on the preparation of patients for laboratory tests 2001 55 63.7 9.1 27.3 2004 78 44.9 19.2 35.9 0.08 b. on the collection and handling of samples 2001 54 59.2 13.0 27.8 2004 79 32.9 25.3 41.8 0.010 c. as consultations by laboratory physicians 2001 31 38.7 35.5 25.9 2004 64 50.0 32.8 17.2 0.50 9. We are satisfied with the physician s reports in a. nuclear medicine 2001 35 68.6 25.7 5.8 2004 48 68.8 22.9 8.4 0.88 b. bone marrow examination 2001 23 78.2 21.7 0.0 2004 40 65.0 30.0 5.0 0.39 c. protein electrophoresis 2001 24 58.4 37.5 4.2 2004 41 61.0 31.7 7.3 0.81 continued 424

Customer satisfaction of laboratory services Table 1 continued Statements assessed Year Number of answers Strongly agree/ agree (%) Neither agree nor Aggregated percentages of the responses and the total number of responses to each statement are given. 1 Significance at the P < 0.05 level. NA, P-values not calculated because of too many low frequencies for chi-squared test. Disagree/ strongly P-value 1... 10. We rarely need to contact the laboratory because of a. missing test results 2001 67 55.2 9.0 35.9 2004 86 61.6 14.0 24.4 0.26 b. erroneous test results 2001 66 78.7 12.1 9.1 2004 86 77.9 14.0 8.1 0.93 11. Turnaround time is adequate for a. stat tests 2001 67 65.7 6.0 28.4 2004 86 67.4 10.5 22.1 0.47 b. routine tests for in-patients 2001 49 63.3 10.2 26.5 2004 68 72.1 8.8 19.1 0.58 c. routine tests for outpatients 2001 38 78.9 5.3 15.8 2004 57 75.4 8.8 15.8 0.81 12. Laboratory has a positive attitude toward our research projects 2001 35 77.2 22.9 0.0 2004 52 80.7 19.2 0.0 NA Table 2 Classification of the most important problems in laboratory services as reported by the clinical units in response to the open-ended question in 2001 Categories The three main categories and their subcategories are shown. 1 In-patient units. 2 Outpatient units. Number of complaints... Laboratory information system (requests and reports) Information to support requesting, reporting, and interpretation is insufficient 15 Requesting of tests is too complicated 13 Electronic requests and reports are not available in some special cases 4 Phlebotomy services Lack of non-scheduled service 1 7 Single defects in phlebotomy skills and customer service attitude 6 Distant location of phlebotomy stations 2 5 Not enough scheduled rounds 1 4 Not enough out-of-hours services for pediatrics 1 4 Long waiting times at phlebotomy stations 2 3 Lack of services for function/tolerance tests 1 2 Delays at the scheduled round at 07:00 hours at weekends 1 1 Test turnaround time Long turnaround time of certain single tests or patient groups in the daytime (mainly clinic of oncology) 11 Long turnaround time (clinic of psychiatry) 6 Long turnaround time because of transportation 4 Long turnaround time during the emergency hours 3 425

P. I. Oja et al. They were also informed about the turnaround times of laboratory tests. Laboratory information system. No corrective actions were carried out for laboratory information system, because it was old and not amenable to major improvement and because renewal of the laboratory information system was planned and accomplished in 2004 05. Turnaround times. The causes for the long turnaround times of tests perceived by the respondents were identified at negotiations. In a typical case, tests were ordered as routine despite emergency needs. These clinical units were advised to order stat tests in urgent cases. In addition, the staff at the clinic of psychiatry had some difficulties in reviewing the reported results in laboratory information system. They were therefore advised on how to use the laboratory information system. The excessive turnaround time due to specimen transportation from the clinic of psychiatry (Table 2) was shortened from 2.5 hours to 30 minutes by synchronizing the transportation process, without changing the number of daily transportation events. Instructions from the laboratory. Instructions for blood collection, including pictures of the tubes used for venipuncture and for pediatric samples, were produced and delivered to the clinical units in 2002 03. The instructions for patient preparation for laboratory tests in clinical chemistry and nuclear medicine were rewritten and delivered to the clinics in 2002 and 2003, respectively. Laboratory Users Handbook was updated in early 2004, paying special attention to the indices and the table of contents. Phlebotomy services. After consulting the in-patient units, the schedules of the main phlebotomy rounds were modified, and most of the skin puncture sampling rounds were combined with the main phlebotomy rounds in 2002. Before this rescheduling, 13 scheduled rounds in 24 hours were available for the clinics. When the rounds were re-scheduled to meet the needs of the clinics, it turned out that the total number of rounds could be reduced to 11 in 24 hours. The medical laboratory technologists working during the emergency hours were given additional practical training on neonatal phlebotomy to meet the needs of out-of-hours phlebotomy services in the clinic of pediatrics (Table 2). Consultations. At the negotiations, the clinicians were asked to specify the kind of consultations with laboratory physicians they needed. It appeared that consultations were needed for the appropriate diagnostic strategies in certain emergency situations, in bleeding disorders, in disorders of amino-acid metabolism, and in endocrinological surgery. Consultations were also needed for better coordination of the laboratory services with the clinical processes. These needs were reported to the laboratory physicians responsible for possible actions. Survey in 2004 Compared with the disagreement levels obtained in 2001, the only significant differences in 2004 were seen in the responses concerning scheduled phlebotomy rounds in the daytime and the instructions on the collection and handling of samples. The dissatisfaction with the phlebotomy rounds decreased from 21.1 to 1.7% (P = 0.005), and the dissatisfaction with the instructions on collection and handling of samples increased from 27.8 to 41.8% (P = 0.010). The respondents who needed additional instructions were contacted by phone, and they were asked to specify their answers. Most of these respondents could not specify their needs. One-fourth of the respondents needed instructions on handling of microbiology specimens. In addition, a few units needed instructions on pretransfusion samples. No significant changes were obtained in the dissatisfaction levels concerning missing test results, laboratory information system, turnaround times, instructions on the preparation of patients, consultations, and the Laboratory Users Handbook, all of which showed high (>20%) levels of dissatisfaction in 2004. The answers of the nurses-incharge showed higher dissatisfaction levels with turnaround times of stat tests than those of the senior physicians. Discussion The purpose of our satisfaction surveys was to identify problems in the laboratory services and to carry out corrective actions. The questionnaire consisted of statements covering different sections and details of laboratory services which were considered important to the laboratory and to the customers. Most important problems were identified by both closedended statements and the open-ended question, supporting the view of representative coverage of essential problems. The questionnaire did not, for example, include statements concerning the reliability of the laboratory tests as a whole. The reliability of laboratory tests is controlled by external and internal quality assessment programs. Incidental errors in laboratory test results are clarified immediately when receiving the complaint. The reproducibility of the results was not specifically pretested in this study with a subgroup of hospital units. However, when comparing the dissatisfaction points in 2001 and 2004 of the topics with no corrective actions after 2001, no changes were observed. This may give some idea of the reproducibility of the measurements. The response rates of the surveys can be considered very good [6]. It is evident that one reason for the very good rates of the clinical units was that two questionnaires were sent to each unit. In addition, our customers were aware that the laboratory has an accredited quality system with organized collection of customer feedback. They may have anticipated that, by giving their comments, laboratory takes notice of the comments for the improvement of its services. The dissatisfaction related to laboratory information system was relatively widespread. This was not unexpected, because it originated from the 1970s. During the 2004 survey, the old system was still in use, causing an even higher dissatisfaction percentage. A new laboratory information system with a graphical user interface was installed in 2004 05 allowing us to improve markedly the requesting of laboratory tests and reviewing the test results. 426

Customer satisfaction of laboratory services The dissatisfaction concerning the instructions on the collection and handling of samples was even higher in 2004 compared with that of 2001. Further clarifications among dissatisfied customers after the 2004 survey showed that a great part of them meant microbiology specimens although our services do not include microbiology. Customers may not differentiate between the specialties of laboratory medicine. It is possible that this has had some effect on the difference in satisfaction between 2001 and 2004. On the other hand, it is somewhat surprising that most of the dissatisfied respondents could not specify which instructions they needed. Increased dissatisfaction rates may also reflect increased awareness of documented instructions, because extensive quality management system documentation has been carried out in our hospital in recent years. Dissatisfaction related to the turnaround time of tests was also high. Many clinical units did not know the real turnaround times of routine laboratory tests. In addition, some units did not use stat requests in urgent cases and thus perceived the results as being delayed. It is also possible that the higher prices of stat tests for the clinics had some effect on the requesting patterns. As a corrective action, the laboratory informed these clinics about the delays of laboratory results and advised them to order stat tests for urgent cases. In many units, the nursesin-charge were more dissatisfied with the turnaround times of stat tests than the senior physicians. This may reflect the workflow of nursing personnel when checking that laboratory results are available for urgent clinical decision making. Some earlier studies have also shown that turnaround time of laboratory tests causes dissatisfaction among hospital physicians. In an American survey of Steindel and Howanitz [10], it was found that 39% of the emergency department physicians did not consider the laboratory sensitive enough to their stat testing needs and turnaround time of the laboratory tests was perceived too long. In a wide survey of Zarbo et al. [11] concerning the satisfaction of physicians with anatomic pathology services, the low satisfaction scores were related to the timeliness of reporting. Furthermore, some other studies have shown that also general practitioners may be dissatisfied with the turnaround times of clinical chemistry laboratory services [12,13]. Phlebotomy rounds were re-scheduled in cooperation with the in-patient units in order to adjust them better with their daily process. It is noteworthy that in addition to the actions of the laboratory, some clinical units revised their own work patterns to integrate them better with the laboratory services. The re-scheduling of rounds resulted in a very significant increase in the satisfaction rate in 2004. As a result of this collaboration, phlebotomy services improved, although the number of rounds decreased. Thus, optimization of laboratory services available to the clinics is a matter of coordinating clinical and laboratory processes. It is evident that clinical laboratories perform customer satisfaction surveys, but the scientific literature concerned is scanty. The modernized quality standards, such as ISO 15189 [1] and ISO 17025 [2], call for the analysis and use of customer feedback and corrective actions, preferably in a systematic way. Comparison of the results of the above-mentioned studies with our study results is difficult, because the specialties of the laboratories varied [11], and the responding customer groups were different [12,13] from those in our study. Also wording of questions and response alternatives varies. Direct comparisons between studies necessitate the use of exactly the same questions [8]. In addition, many customer surveys are aimed to local needs and thus special questionnaires must often be designed to address local problems for improvement [14]. The laboratory process starts from the request for a test made in the clinic and ends when the clinician receives the test result. In efficient cooperation, the clinical and laboratory processes are mutually compatible. This means that both parties work in constant interaction. Different forms of interaction at the laboratory s customer interface, e.g. satisfaction surveys, comment cards, discussions, meetings, and site visits, make the parties more aware of the activities of the other. It is essential for the laboratory to be familiar with the customers workflow practices, and conversely, clinical customers need to know laboratory processes. Often changes in the practices of interactions needed may take years. The final assessment of laboratory processes must, nevertheless, be based on the laboratory s expertise. If a laboratory makes changes based on customers preferences only, the outcome may be decreased efficiency and increased cost of the laboratory services. Customer surveys give information of the overall satisfaction with the aspects of interest. With additional contacts, we were able to find out the specific problems of the clinical units for accurate corrective actions. It is, however, natural that some of the corrective actions may be carried out immediately, and others must wait for an appropriate time for, e.g., investments. In line with our quality system, we will continue regular satisfaction surveys at the clinics by using the cycle of continuous quality improvement: survey data analysis planning and performing corrective actions survey. References 1. ISO 15189:2003. Medical laboratories Particular requirement for quality and competence. 2. ISO/IEC 17025:2005. General requirement for the competence of testing and calibration laboratories. 3. Kaplan RS, Norton DP. Translating Strategy into Action. The Balanced Scorecard. Boston: Harvard Business School Press, 1996. 4. Kelley K, Clark B, Brown V, Sitzia J. Good practice in the conduct and reporting of survey research. Int J Qual Health Care 2003; 15: 261 266. 5. Streiner DL, Norman GR. Health Measurements Scales. A Practical Guide to Their Development and Use, third edition. Oxford: Oxford University Press, 2003. 6. Mangione TW. Mail surveys. In Bickman L, Rog DJ, eds. Handbook of Applied Social Research Methods. Thousand Oaks: Sage Publications, 1998: 399 427. 7. Fowler FJ Jr. Design and evaluation of survey questions. In Bickman L, Rog DJ, eds. Handbook of Applied Social Research Methods. Thousand Oaks: Sage Publications, 1998: 343 374. 427

P. I. Oja et al. 8. Bourque LB, Fielder EP. How to Conduct Self-Administered and Mail Surveys. Thousand Oaks: Sage Publications, 1995. 9. Weber PR. Basic Content Analysis, second edition (Sage University Paper Series on Quantitative Applications in the Social Sciences, 07-049). California: Sage Publications, 1990. 10. Steindel SJ, Howanitz PJ. Physician satisfaction and emergency department laboratory test turnaround time. Observations based on College of American Pathologists Q-Probes studies. Arch Pathol Lab Med 2001; 125: 863 871. 11. Zarbo RJ, Nakhleh RE, Walsh M. Customer satisfaction in anatomic pathology. A College of American Pathologists Q-Probes study of 3065 physician surveys from 94 laboratories. Arch Pathol Lab Med 2003; 127: 23 29. 12. Allen KR, Harris CM. Measure of satisfaction of general practitioners with the chemical pathology services in Leeds Western Health District. Ann Clin Biochem 1992; 29: 331 336. 13. Boyde AM, Earl R, Fardell S, Yeo N, Burrin JM, Price CP. Lessons for the laboratory from a general practitioner survey. J Clin Pathol 1997; 50: 283 287. 14. Zarbo RJ. Determining customer satisfaction in anatomic pathology. Arch Pathol Lab Med 2006; 130: 645 649. Accepted for publication 5 September 2006 428