QJM Advance Access originally published online on August 26, 2005
QJM 2005 98(10):753-756; doi:10.1093/qjmed/hci116
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Near-miss errors in laboratory blood test requests by interns
From the Departments of 1Medicine & Therapeutics and 2Chemical Pathology, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong, SAR, China
Address correspondence to Dr K.M. Chow, Department of Medicine and Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong. email: chow_kai_ming{at}alumni.cuhk.net
Received 11 May 2005 and in revised form 13 July 2005
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Background: Human errors have proven to be one of the most formidable patient care challenges in acute hospital setting.
Aim: To evaluate the at-risk period for near-miss errors in laboratory blood test requests, in an acute medical hospital.
Design: Hospital-based retrospective analysis.
Methods: We reviewed the database of voluntary reports for near-miss errors for laboratory blood test requests by 104 medical residents in their first postgraduate year (interns), over a 2-year period (October 2002 to September 2004). To identify patterns and causal factors we analysed the reports with respect to months of working experience, work hours, and work shifts of an extended duration.
Results: There were 52 near-miss events among patients cared for by the medical service (20 male patients, 32 females, mean age 72.6 ± 9.7 years). The overall incidence of near-miss events when interns practiced during the first month of training vs. subsequent months was 1.6 (95%CI 0.77–2.9) vs. 0.6 (95%CI 0.44–0.83) cases per 100 intern-days at risk. The odds ratio for a near-miss event during the first month of intern training vs. subsequent months was 2.64 (95%CI 1.29–5.38). With respect to the interns’ on-call shift schedule, one half of the near-miss episodes occurred during an intern's on-call days and another half of them during an extended on-call shift; none of the events occurred during a standard working shift. These events peaked in frequency when on-call interns had worked for 12–20 h.
Discussion: The first month of internship represents an error-prone period. The best interventions to reduce near-miss errors by recently graduated medical interns should be the subject of further research.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
After years of discussion, the pressure to reduce medical errors has remained top of the agenda in patient care improvement, and the root causes of medical errors continue to be actively sought.1,2 To address the question of whether there exists a susceptible period for medical interns to make errors during their training program in acute hospitals, we reviewed all near-miss medical errors in blood test ordering for patients admitted to the medical service over a 2-year period. Such information is particularly important because it might represent a modifiable or predictable risk factor, or alternatively, offer a safety net for early identification of medical incidents.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
In our tertiary-care urban teaching hospital with over 300 acute medical beds in Hong Kong, medical patients are cared for by attending physicians and a team of 13 interns who have average weekly work hours of 80 h. In other words, the interns have overnight on-call duty every fourth night. Consecutive hours of work are limited to 30, with an average of 1 day off per week.3
Only patients under the care of medical departments were eligible for this study and included those under the general medical, cardiology, endocrinology, gastroenterology, geriatrics, haematology, respiratory, nephrology, neurology and rheumatology services; intensive care unit patients were not included. Near-miss medical errors in laboratory blood test requests by the interns were identified by previously described self-report methods.4 Confidential reporting of near-miss events, in general, was made by the medical interns within 24 h of occurrence to the on-duty pathologist, using an email system or other internal communication means. The reporting mechanism was primarily derived to capture wrong test results due to human error, and invalidate the misleading laboratory results. In other words, the near-miss events refer to wrong blood samples for laboratory investigations such as biochemistry panel and blood gas analysis. We focus on preventable or procedural errors: specimen samples being mislabelled with another patient's identification; blood ordered for the wrong patient; and a blood sample being taken from the wrong patient. All reported events were reviewed and confirmed by the on-duty pathologist and independent investigators.
The computerized blood test ordering allowed us to determine precisely when all these near-miss events occurred. Once the near-miss events were identified, additional information was documented, including the time and date of the order, the work schedule, working experience and work shifts of an extended duration for the involved intern. The interns’ working day was divided into three working schedules in rotation, namely on-call days, extended on-call shift, and standard working shift.
All data are reported as means ± SD unless otherwise specified. Estimates of incidence are presented as 95%CIs, assuming the data are in Poisson distribution. We analysed the rates of near-miss laboratory blood test request errors with reference to the interns’ work schedule, and further correlated the incidence for each month to their months of working experience. The event rate in our study was represented by the number of near-miss errors divided by the total number of opportunities for errors. An opportunity for error was defined as number of 100 intern-days at risk. By comparing the ‘expected’ (calculated from the overall event rate) and the observed number of incidents recorded for each calendar month, using 
2 goodness-of-fit test, we tested the assumption that the incidence of near-miss procedural errors would be constant throughout the study period. To determine further the extent of variability in the incidence of near-miss events, we used the same test to directly compare event rates for the months associated with the highest values relative to the overall incidence observed. The comparative risk of intern-associated medical errors, stated as the medical error events per intern-day, was then evaluated according to for each of the following three work schedules: on-call days, extended on-call shift, and standard working shift. Statistical significance was defined as p = 0.05. The number of work hours at the time of the near-miss error was also identified and correlated.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
From October 2002 to September 2004, 52 near-miss cases occurred among patients cared for on the medical service. Among these 52 (20 males, 32 females), the mean patient age was 72.6 ± 9.7 years. Most cases (71%) occurred because specimen samples were mislabelled with another patient's identification, whereas a blood sample was taken from the wrong patient in 23%. As suggested by the term ‘near-miss’, the errors were detected before harm occurred to patients.
There was no significant variation in the rate of near-miss events across the week (overall 2 test, p = 0.98). Nineteen (39.6%) episodes occurred outside normal office hours (9 am to 5 pm), although data for the exact timing of these events, defined as the time of blood-taking, were missing in eight cases. There was a trend of monthly variation in the near-miss event rate (overall 2 test, p = 0.057). However, there was a significant increase in near-miss wrong blood taking events in July, when the new graduate medical interns start the training program (Figure 1). The overall incidence of near-miss events when the interns practiced during the first month of training vs. subsequent months was 1.6 (95%CI 0.77–2.9) vs. 0.6 (95%CI 0.44–0.83) cases per 100 intern-days at risk. The odds ratio for near-miss events during the first month of training vs. in subsequent months was 2.64 (95%CI 1.29–5.38).
|
One half of the near-miss episodes occurred during on-call days, and the other half during an extended on-call shift; none occurred during a standard working shift. The event rate during on-call days was 0.47 (95%CI 0.28–0.72) cases per 100 intern-days at risk, similar to the event rate during extended on-call shift of 0.47 (95%CI 0.28–0.72) cases per 100 intern-days at risk. The 95%CI for event rate during standard shift was 0–0.15 cases per 100 intern-days at risk (Figure 2). These event rates peaked when the on-call interns had worked for 12–20 h (Figure 3).
|
|
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
In this retrospective hospital-based study, the odds that interns had a near-miss event in the first month of training were more than double the odds in the subsequent 11 months. Near-miss incidents were also more likely to occur during on-call days and extended on-call shifts. These findings, which are of particular concern in today's climate of pressure to reduce medical errors, are consistent with the premise that patient safety can be addressed with improved residency training programs and working schedules.
Our data on the high incidence of near-miss medical errors among recently graduated medical interns in the first month of rotation indicate that inexperience and unfamiliarity with hospital-specific systems could also be implicated. Although conclusion from previous studies on the effect of the ‘July phenomenon’5–9 have been mixed, our findings add to these earlier studies in important ways. First, this study involved a cohort of medical interns with working hours limited to 80 h per week, now regarded as the accepted standard in most countries, as well as by the Accreditation Council for Graduate Medical Education in the US.3,10 Furthermore, we analysed the incidence of near-miss errors, which might be more reflective of intern performance. For example, although most studies across a variety of specialties did not demonstrate a ‘July phenomenon’ on hard outcomes such as mortality,7,11 it is quite likely that closer supervision and extra vigilance by the senior staff compensate for the relative inexperience of interns. Stated another way, retrospective analyses on adverse outcomes (instead of near-miss events) tend to underestimate the true incidence of medical errors, simply because near-miss events do not result in adverse outcomes. In fact, when near-miss errors or more representative indicators of medical performance such as ordering errors were measured, the July phenomenon or effect of the intern experience was often significant.5,8,12,13 As such, instead of addressing whether a July phenomenon exists, we believe that one of the major implications from our study is to highlight the relevance of planned close supervision during the error-prone period.
As in previous studies,9,14 most of the medical errors were made by on-call interns. One explanation is that on-call interns get most of the calls from the medical wards, make more medical decision and carry out more procedures that are prone to errors. Nevertheless, the effect of call status on the rate of erroneous blood test ordering should ideally be correlated with the denominator of case loads (the number of blood tests being carried out during the on-call shift), which unfortunately was not available in our study.
This study also has other limitations. It reflects the experience in only one large teaching hospital, and the results may not extrapolate to other institutions. We analysed only one element of an intern's work-related activities. It remains uncertain whether intern experience or work shift would influence the frequency of other avoidable events such as diagnostic and medication errors. Finally, we are unable to determine the exact mechanism of increased near-miss events during the first month of internship.
In conclusion, the first month of medical intern training and on-call shifts were two unique periods susceptible to high rates of near-miss errors within a 2-year observation period in an acute medical unit of teaching hospital. Our results may have important implications for health policy. Further research is warranted to evaluate whether undergraduate training or internship supervision would be the best contributors to improved patient care.
|
Acknowledgments |
|---|
We are indebted to Ms Linda Leung and Ms Margaret Wong for clerical support.
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Davydov L, Caliendo G, Mehl B, Smith LG. Investigation of correlation between house-staff work hours and prescribing errors. Am J Health Syst Pharm 2004; 61:1130–4.
2. Battles JB, Shea CE. A system of analyzing medical errors to improve GME curricula and programs. Acad Med 2001; 76:125–33.[ISI][Medline]
3. Steinbrook R. The debate over residents' work hours. N Engl J Med 2002; 347:1296–302.
4. O'Neil AC, Petersen LA, Cook EF, Bates DW, Lee TH, Brennan TA. Physician reporting compared with medical record review to identify adverse medical events. Ann Intern Med 1993; 119:370–6.
5. Shulkin DJ. The July phenomenon revisited: are hospital complications associated with new house staff? Am J Med Qual 1995; 10:14–17.
6. Myles TD. Is there an obstetric July phenomenon? Obstet Gynecol 2003; 102:1080–4.
7. Barry WA, Rosenthal GE. Is there a July phenomenon? The effect of July admission on intensive care mortality and length of stay in teaching hospitals. J Gen Intern Med 2003; 18:639–45.[CrossRef][ISI][Medline]
8. Walling HW, Veremakis C. Ordering errors by first-year residents: evidence of learning from mistakes. Mo Med 2004; 101:128–31.[Medline]
9. Borenstein SH, Choi M, Gerstle JT, Langer JC. Errors and adverse outcomes on a surgical service: what is the role of residents? J Surg Res 2004; 122:162–6.[CrossRef][ISI][Medline]
10. Philibert I, Friedmann P, Williams WT; the ACGME Working Group on Resident Duty Hours. New requirements for resident duty hours. JAMA 2002; 288:1112–14.
11. Buchwald D, Komaroff AL, Cook EF, Epstein AM. Indirect costs for medical education: is there a July phenomenon? Arch Intern Med 1989; 149:765–8.[Abstract]
12. Rich EC, Gifford G, Luxenberg M, Dowd B. The relationship of house staff experience to the cost and quality of inpatient care. JAMA 1990; 263:953–7.[Abstract]
13. Rich EC, Hillson SD, Dowd B, Morris N. Specialty differences in the ‘July Phenomenon’ for Twin Cities teaching hospitals. Med Care 1993; 31:73–83.[CrossRef][ISI][Medline]
14. Veasey S, Rosen R, Barzansky B, Rosen I, Owens J. Sleep loss and fatigue in residency training: a reappraisal. JAMA 2002; 288:1116–24.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||