Strategies for improving the collection of 24-hour urine for analysis in the clinical laboratory: redesigned instructions, opinion surveys, and application of reference change value to micturition.
To streamline the analytic care of patients and avoid delays in providing health care, decentralization of the tasks required to obtain clinical specimens has become widespread in modern medical facilities. This decentralization means that the laboratory must enhance its supervision duties to prevent errors that would reduce the quality of analytic results or of the service provided. (1)
Common sources of error for nonlaboratory personnel include inadequate training of nursing staff in specimen collection techniques, lack of familiarity about the types of determinations performed in the laboratory, incorrect use of additives and devices for each determination, incorrect maintenance of temperature conditions, and delays related to the maximum period of time for submitting samples to the laboratory. All of these factors are primary sources of laboratory errors. (2)
In the case of 24-hour urine sample collection, the problems are exacerbated because active participation of the patients is necessary. Most laboratories provide specific written instructions to reinforce the verbal explanations given to the patient and provide suitable containers for each type of determination to be conducted. These instructions should be clear, simple, and easily understood by all patients.
The reality is that, often, in our laboratory, (3) patients do not give the 24-hour urine samples requested by the clinician to the staff in the hospital laboratory who know the protocol to proceed with their handling. On other occasions, the samples are not collected under the right conditions or micturition is incomplete. Moreover, these findings are common to many laboratories in our setting. (4)
On the other hand, satisfaction surveys, together with the opinions of internal staff members (doctors and nurses, end-users of the analytic results) or external "customers" (patients), are considered quality elements and are even deemed to be essential by those in charge of monitoring the quality control systems implemented in clinical laboratories. (5,6) The results of these surveys and the opinions expressed are important tools for ongoing improvement of the laboratory. They provide information on the quality of the services rendered by the laboratory and include suggestions made by the interviewees for implementing improvement strategies. (7,8,9)
The aim of our study was to carry out an intervention among nonhospitalized patients (outpatients from external hospital surgery services and from public health centers) to determine whether the distribution of adequate information to staff responsible for collecting the 24-hour urine specimens, who then provide this information to patients, results in fewer preanalytic incidents. Furthermore, patient opinion surveys on the quality of information material were conducted. Based on these data, specific improvements may be proposed.
Before the intervention, preanalytic incidents during urine collection were evaluated for a 1-month period and reported by laboratory personnel. One year later, after the implementation of the new recommendations, the number of preanalytic incidents was reevaluated.
In addition, our aim was to capitalize on the micturition data of patients with preintervention incidents to determine the intraindividual biologic variability and thus predict the proper, complete collection of 24-hour urine for each patient on the basis of their usual rate of micturition. There are many references in the scientific literature dealing with the application in clinical practice of the data on biologic variation. One of the most interesting, the reference change variable (RCV), evaluates the meaning of the changes in the serial results for a particular individual, for each parameter. The concept of RCV was developed by Harris (10,11) and Yasaka (11) and defined as that difference between 2 consecutive test results for an individual that is statistically significant for a given proportion of persons all with similar characteristics. By allowing for variation in within-person variances, this procedure computes a reference change that is more specific than that obtained directly from the distribution of observed differences between measurements. The essential idea is to select as a reference change that difference between consecutive values that would be statistically significant (eg, P < .05) for most individuals concerned (95% of such persons). This requires taking into account within-person variances under steady-state conditions and the distribution of such variances across the selected population.
In our study, we determined the value of the intraindividual biologic variability of the participating patients and estimated the RCV of 24-hour micturition in a large sample of the population. Using these intraindividual variability values in our setting, we analyzed the variability in patients included in the study during the opinion survey collection period, together with any possible associated variables. We also studied the variability in patients who reported preanalytic incidents in the postintervention period.
MATERIALS AND METHODS
The intervention took place in the 20th health district of the Valencian community in Spain, between November 2006 and December 2007, and was coordinated by the Clinical Laboratory Department of the University Hospital of Elche (Elche, Spain). The intervention was approved by the managing director of the health district and by the hospital's research committee.
Estimation of the preanalytic incidents related to 24-hour micturition before laboratory intervention was quantified by using the incidents register of the hospital laboratory's specimen reception area between October 11 and November 11, 2006. The variables recorded were the following: total number of 24-hour urine requests received, total number of samples eventually accepted by the laboratory (number of 24-hour urine and acidified 24-hour urine samples), and the total number of incidents recorded (no micturition details, sample not received in laboratory or rejected because it did not comply with established collection conditions). During the same period of time the following year, incidents were evaluated after the intervention (from October 11, 2007, to
November 23, 2007).
To homogenize the information given by the laboratory staff to the health care staff, which in turn was given to patients, all the instructions necessary to obtain the requested 24-hour urine samples were reviewed and/or drawn up, together with the preanalytic conditions the patients had to follow for proper urine collection. The formats were homogenized and distributed among health care personnel of the health district. Meetings were organized with the staff who are in contact with the patients in both the hospital outpatient department and the public health centers to coordinate the tasks of providing patients with information.
An opinion survey was conducted among all patients of the 20th health district of the Valencian community who were seen in primary care and specialized outpatient clinics between December 2006 and July 2007 and whose laboratory test request forms included determinations in 24-hour urine. This survey was performed to complete the data integrated in the recommendations and to reinforce the information given to health care personnel on key aspects. The opinion survey on the 24-hour urine collection process was given consecutively to all patients, together with the necessary containers--with or without additives--and corresponding instructions. The survey was designed to obtain opinions on the problems previously detected and related to 24-hour urine collection. (3) The survey was drawn up based on the information received by the staff in direct contact with the patients at the appointment desk and the staff in the urine specimen reception area (Figure 1).
The data relating to prior 24-hour micturition of the patients in the study was obtained from existing registers in the laboratory's database (OpenLab, Nexus IT, Las Rozas, Spain).
A descriptive study of the variables obtained from the surveys was done, and the differences between preintervention and postintervention preanalytic incidents were compared using the [chi square] test.
The RCV was determined by using the Harris formula10:RCV = 2 1/2 x Z x [[CV.sup.2.sub.A] + [[CV.sup.2.sub.I]].sup.1/2], where [CV.sub.A] is the coefficient of analytic variation for the method of measuring 24-hour micturition; [CV.sub.I] corresponds to the intraindividual variation in 24-hour volume; and P < .05, Z = 1.96.
The lack of [CV.sub.I] data for the evaluation of 24-hour micturition, and the need for its inclusion in the Harris formula, led us to propose the inclusion of a new parameter in the formula, the percentage of variation in volume of 24-hour urine or PVVI.
PVVI = [(Current 24-hour Micturition - Historical 24-hour Micturition) / (Current 24-hour Micturition)] x 100
As shown, PVVI is based on the biologic variation of patients' micturition.
[FIGURE 1 OMITTED]
The variables studied in the patient surveys and the RCV were compared by using a single-factor analysis of variance (ANOVA) for continuous variables and the [chi square] test for categorical variables.
Results of the Surveys During the Study Period
A total of 302 patients completed the opinion survey between December 2006 and July 2007.
The results corresponding to demographic and organizational variables are shown in Table 1. Most of the urine samples received came from specialized care services (244, 80.8%) and most were handed in at the hospital (219, 72.5%). The mean period of time elapsed between provision of information and submission of the urine specimen was 38.8 days (standard deviation, 52.7). Nephrology was the service that most frequently requested 24-hour urine samples (125, 41.4%), followed by internal medicine (65, 21.5%) and family or general medicine (55, 18.2%). Table 2 shows the variables relating to the patients' opinion concerning the collection of 24-hour urine specimens. More than half of the patients had not submitted this type of sample to the laboratory before (194, 64.2%). Most received instructions from the relevant health care personnel verbally (266, 88.1%) or in written form (253, 83.8%), and in most cases they understood them perfectly (271, 89.7%). Regarding the difficulties in collecting the sample, 40 of 302 patients (13.2%) surveyed reported 97 problems in all. These complications were related to non-fulfillment of the collection schedule (36, 37.1%), loss of urine specimen (25, 25.8%), and insufficient number of containers provided for collection (24, 24.7%) and to for getting the instructions (12, 12.4%). In addition, more than half of the patients had to collect various urine samples sequentially (175, 57.9%). Only 6 patients (2.9%) filled in the section for "Comments and suggestions" and these comments referred to the same problems investigated in the survey.
Comparison of Preanalytic Incidents Before and After the Intervention
During the preanalytic incidents study period, a total of 1127 requests were made for 24-hour urine determinations, of which 997 (88.5%) were accepted by the laboratory. During the study period after the intervention, 883 requests were made, of which 807 (91.4%) were accepted by the laboratory (P = .03) (Table 3). Before the intervention, 130 preanalytic incidents were recorded (11.5%), and after the intervention, 76 (8.6%) (P = .04) were recorded. Of the 130 incidents recorded before the intervention, 66 (50.8%) corresponded to requests that were not accompanied by a specimen and 63 (48.5%) to requests in which the micturition volume was not recorded. In comparison, most of the incidents recorded in the postintervention period corresponded to requests not accompanied by a sample (73, 96.1%), with only 1 (1.3%) incident in which the micturition volume was not recorded (P < .001 for both comparisons).
Analysis of the RCV and its Relationship With the Variables Studied
Determination of RCV.--To calculate the RCV in our setting, we used the patients' data to collect information on the preintervention incidents. During this period, a total of 997 urine samples were submitted to the laboratory. Of these, 918 (92.1%) were included in the study because data on prior micturition were available; thus, we could use these to calculate PVVI.
In this series, PVVI was 19.0 [+ or -] 16.5% (mean [+ or -] standard deviation). Figure 2 shows the dispersion of PVVI values for 24-hour micturition for the patients in the study. For a coefficient of analytic variation ([CV.sub.A]) of 5% for the micturition measurement, the RCV was 54.5%. This means that a patient's micturition associated with a PVVI value outside the +54.5% range suggests that 24-hour urine collection was incomplete.
Taking this RCV value as a reference, we describe, on the one hand, the variation among patients in the study for whom 24-hour micturition data were available (patients belonging to the group that filled in the opinion survey and patients for whom preanalytic incidents were recorded after the intervention) and on the other, we describe the variables associated with these variations that are based on the results of the opinion survey.
Patient Variability Based on PVVI--Of the 302 patients who completed the opinion survey, historical micturition data were available for 201 (66.6%). In addition, data on prior micturition were also available for 436 of 807 patients (54.0%) who submitted samples to the laboratory during the postintervention period. Therefore, 637 patients were included in the analysis of variability. Using the patients' PVVI values and comparing them with the RCV previously determined, we calculated that a total of 597 patients (93.7%) had a PVVI value inside the [+ or -] 54.5% range (mean, 17.7; SD = 14.5); however, 40 patients (6.3%) had a PVVI value outside the [+ or -] 54.5% range (mean, 80.1; SD = 23.5).
Analysis of Variables Related to the Variation in
PVVI.--Table 4 shows the variables studied in the opinion surveys of 201 patients, with historical micturition data as a function of the variation in PVVI in accordance with the RCV range determined in our population ( [+ or -] 54.5%). Older patients had a greater tendency to present urine samples with PVVI values outside the range (P = .02). We also observed differences in relation to the service requesting the samples. The largest percentage of patients with PVVI values within the range determined for our population were patients from general medicine. Patients from services other than nephrology, internal medicine, and general medicine had a higher percentage of a PVVI value outside the +54.5% range. There were no significant differences in the other variables studied.
The results of our study have enabled us to gain a greater insight into the origin of the problems involved in the collection of 24-hour urine from patients in our health department.
The opinion surveys show that the measures taken by the laboratory to inform patients on proper collection procedure were sufficient because more than 83% of the patients received verbal or written instructions that they understood. Problems were detected in 13.2% of cases and may be related to the complexity involved in collecting various urine samples within a short period of time (reported by 57.9% of the patients surveyed). Other problems detected related to forgetting instructions in 12.4% of cases and nonfulfillment of collection schedule in 37.1%.
Although we started with a number of preanalytic incidents that is not very high (130/1127, 11.5%), similar to data published previously by one of us (C.T.) (3) and by other authors, (4) the results show that the intervention has resulted in an improvement (76/883, 8.6%). While it is true that the number of incidents related to failure to submit the sample to the laboratory did increase, the number of 24-hour urine samples without a volume indication markedly decreased.
The RCV has proved to be a useful tool for interpreting the results of the laboratory tests by using longitudinal comparisons. This tool enables decisions to be taken concerning the importance of the changes observed in serial results for a particular patient. (12) Databases on 261 analytes have been published, (13) but they do not include data on patients' micturition. Although micturition is measured and reported by the laboratory, it is not considered an analytic biologic variable, such as the measurement of proteins or glucose in urine, because calibration or controls are not used. However, in our laboratory, we have evaluated the interobserver variability while analyzing the 24-hour micturition through visual inspection and have obtained a variability of 5%. Therefore, we think that the use of CVA and, hence, RCV in these analysis could improve the quality of the results. Thus, despite the fact that preanalytic aspects are not included in the formula for calculating RCV, which is a serious limitation in its application, (14) there is no doubt that its application to micturition may affect the quality of the analytes measured in 24-hour urine. Another specific problem with micturition is the lack of published data on the individual biologic variability for this variable, which is the reason we propose the use of PVVI instead of CVI.
Application of RCV to our group of patients showed that in only 6.3% of cases should the 24-hour urine samples have been rejected as inadequate.
Finally, analysis of the relationship between the variables studied with the opinion survey and application of RCV value to PVVI indicates that there is a significant association between age and proper collection of 24-hour urine (P = .02).
The intervention resulted in an improvement in quality of urine samples submitted to the laboratory and a decrease in the percentage of preanalytic incidents. The RCV is an objective index for determining whether the 24-hour urine sample has been collected properly.
Measures taken by the laboratory resulted in better communication and outcomes for all personnel involved in analytic and extra-analytic activities. Thus, although the instructions given to collect 24-hour urine are adequate, we propose that laboratory staff verify that the samples were properly collected before carrying out the analysis.
This study received financial support from the Quality and Patient Care General Management of the Valencia Government Department of Health (Spain) by decision of the general director of the Valencia Health Agency on October 20, 2006.
(1.) Salve Martinez ML, Sempere Molina C. Fase premetrologica: efecto del transporte de especimenes de orina sobre los resultados [in Spanish]. Quim Clin. 1995;14(1):12-16.
(2.) Tormo Diaz C, Yago Galvan C, Cebria Artiga J. Implementacioon de un Sistema de Calidad en un laboratorio clinico: revision del estado actual del tema y propuesta de un plan estrategico de establecimiento [in Spanish]. Quim Clin. 1996;15(4):206-211.
(3.) Tormo C. Estimacion del grado de error preanalftico en el area de recepcion de muestras del laboratorio clinico [in Spanish]. Anal Clin. 1997;21:63-69.
(4.) Prieto S, Sempere C, Salve ML, Moreno JM, Grupos de trabajo de AEBM. Metodologia para la estimacion del error preanalitico y su significacion en determinaciones realizadas a partir de especimenes obtenidos en puntos perifericos de obtencion y recogida de especimenes (PPORE) [in Spanish]. Rev Diag Biol. 2003;52(1):46-54.
(5.) Laboratorios Clinicos: Requisitos Particulares Relativos a la Calidad y la Competencia. Madrid, Spain: AENOR; 2003. UNE-EN ISO 15189.
(6.) Sistemas de Gestion de la Calidad: Requisitos. Madrid, Spain: AENOR; 2000. UNE-EN ISO 9001.
(7.) Zarbo RJ, Nakhleh RE, Walsh M. Customer satisfaction in anatomic pathology. Arch Pathol Lab Med. 2003;127:23-29.
(8.) Zarbo RJ. Determining customer satisfaction in anatomic pathology. Arch Pathol Lab Med. 2006;130:645-649.
(9.) Jones BA, Walsh MK, Ruby SG. Hospital nursing satisfaction with clinical laboratory services: a College of American Pathologists Q-Probes study of 162 institutions. Arch Pathol Lab Med. 2006;130(12):1756-1761.
(10.) Harris EK. Some theory of reference values: comparison of some statistical models of intraindividual variation in blood constituents. Clin Chem. 1976;22: 1343-1350.
(11.) Harris EK, Yasaka T. On the calculation of a "reference change" for comparing two consecutive measurements. Clin Chem. 1983;29:25-30.
(12.) Ricos C, Domenech MV, Perich C. Analytical quality specifications for common reference intervals. Clin Chem Lab Med. 2004;42(7):858-862.
(13.) Ricos C, Cava F, Garcia-Lario JV, et al. The reference change value: a proposal to interpret laboratory reports in serial testing based on biological variation. Scand J Clin Lab Invest. 2004;64(3):175-184.
(14.) Omar F, van der Watt GF, Pillay TS. Reference change values--how useful are they? [published online ahead of print March 6, 2008]. J Clin Pathol. 2008; 61:426-42 7. doi:10.1136/jcp.2007.054833.
Consuelo Tormo, PhD; Blanca Lumbreras, PhD; Ana Santos, ChPh; Luis Romero, ChB; Minerva Conca, BSN
Accepted for publication February 20, 2009.
From the Clinical Laboratory Department, Hospital Universitario de Elche, Elche, Alicante, Spain (Dr Tormo, Mr Romero, and Mss Santo and Conca); and Departamento de Salud Publica, Historia de la Ciencia y Ginecologia, Universidad Miguel Hernandez, Campus de San Juan, San Juan, Alicante, Spain (Dr Lumbreras).
The authors have no relevant financial interest in the products or companies described in this article.
Presented in part as a poster at the 29th Aspher Annual Conference '07, Valencia, Spain, October 27, 2007.
Reprints: Consuelo Tormo, PhD, Departamento de Laboratorio Clinico, Hospital Universitario de Elche, Caml de L'Almazara 11, 03203 Elche, Alicante, Spain (e-mail: firstname.lastname@example.org).
Table 1. Demographic and Organizational Data From Surveys Study Characteristic Variable No. (%) Mean (SD) Sex Male 150 (49.7) Female 152 (50.3) Age y 58.9 (15.7) Setting Specialized care 244 (80.8) Primary care 58 (19.2) Urine specimen reception center Hospital 219 (72.5) Health center 83 (27.5) Delay, d 38.8 (52.7) Requesting service Nephrology 125 (41.4) Internal medicine 65 (21.5) General medicine 55 (18.2) Endocrinology 15 (5.0) Oncology 12 (4.0) Rheumatology 8 (2.6) Gastroenterology Urology 4 (1.3) Others 10 (3.3) Table 2. Patient Opinions of 24-Hour Urine Collection Process "Yes" "No" Response, Response, Variable No. (%) No. (%) Prior experience with 24-hour urine collection 108 (35.8) 194 (64.2) Patients given verbal instructions 266 (88.1) 36 (11.9) Patients given written instructions 253 (83.8) 49 (16.2) Patients attesting to understanding the instructions 271 (89.7) 28 (9.3) Problems during sample collection 97 (32.1) 205 (67.9) Nonfulfillment of collection schedule 36 (37.1) Loss of urine sample 25 (25.8) Insufficient containers for urine collection 24 (24.7) Instructions forgotten 12 (12.4) Various urine samples collected sequentially 175 (57.9) 127 (42.1) Table 3. Comparison of Preanalytic Incidents Before and After Intervention Collection Data, No. (%) Variable Year 2006 Year 2007 P (a) Total no. of 24-hour urine requests 1 127 (100.0) 883 (100.0) Total no. of 24-hour urine samples submitted 997 (88.5) 807 (91.4) .03 Total no. of incidents 130 (11.5) 76 (8.6) .04 Total no. of incidents (distribution) 130 76 No urine sample 66 (50.8) 73 (96.1) <.001 24-hour urine sample with no volume indicated 63 (48.5) 1 (1.3) <.001 Urine sample rejected 1 (0.7) 2 (2.6) .64 (a) [chi square] test. Table 4. Relationship Between Percentage of Variation in Volume of 24-Hour Urine (PVVI) and Variables in the Opinion Survey PVVI Inside the Outside the Variable 54.5% Range 54.5% Range P (a) Age, y (mean, SD) 53.0 (17.1) 61.4 (13.9) .02 (b) Sex, No. (%) .74 Male 88 (48.1) 7 (43.8) Female 95 (51.9) 9 (56.3) Setting, No. (%) .40 Specialized care 144 (78.7) 14 (87.5) Primary care 39 (21.3) 2 (12.5) Urine specimen reception center, .13 No. (%) Hospital 127 (69.4) 14 (87.5) Health 56 (30.6) 2 (12.5) Delay, d (mean, SD) 38.9 (48.7) 39.4 (61.9) .98 Service, No. (%) .02 Nephrology 38 (20.2) 2 (11.1) Internal medicine 35 (18.6) 1 (5.6) General medicine 79 (42.0) 6 (33.3) Others (c) 36 (19.1) 9 (50.0) Prior experience in 24-hour .91 urine collection, No. (%) Yes 66 (36.1) 6 (37.5) No 117 (63.9) 10 (62.5) Patients given verbal .96 instructions, No. (%) Yes 161 (88.0) 14 (87.5) No 22 (12.0) 2 (12.5) Patients given written .31 instructions, No. (%) Yes 28 (15.3) 4 (25.0) No 155 (84.7) 12 (75.0) Patients attesting to .86 understanding the instructions, No. (%) Totally 17 (9.4) 1 (6.7) Partially 162 (89.5) 14 (93.3) Patients reporting difficulties 24 (96.0) 1 (4.0) .37 in collecting sample, No. (%) Various urine samples collected .38 sequentially, No. (%) Yes 112 (61.2) 8 (50.0) No 71 (38.8) 8 (50.0) (a) [chi square] test. (b) Single-factor ANOVA test. (c) Endocrinology, oncology, rheumatology, gastroenterology, urology services, etc.
|Printer friendly Cite/link Email Feedback|
|Author:||Tormo, Consuelo; Lumbreras, Blanca; Santos, Ana; Romero, Luis; Conca, Minerva|
|Publication:||Archives of Pathology & Laboratory Medicine|
|Date:||Dec 1, 2009|
|Previous Article:||Whole-slide imaging digital pathology as a platform for teleconsultation: a pilot study using paired subspecialist correlations.|
|Next Article:||Adequate histologic sampling of breast magnetic resonance imaging--guided core needle biopsy.|