Delays in reporting critical values from clinical laboratories to responsible healthcare staff

Objective: We aimed to investigate the frequency of delayed notifications and probable causes of delays for critical value notification in clinical laboratory of university hospital. Materials and methods: All data was obtained from critical value reporting forms and laboratory information system. The frequency and location of critical and delayed results, latencies throughout a working day and the pro-fessional status who received the critical callbacks were shown as percentages. Results: A total of 2018 (1.02%) critical values were reported and 13.1% of them were delayed notifications. Most of them were observed in laboratory tests ordered from patients of service and polyclinics compared to ICU and emergency department (26.7%, 26%, 6.2% and 4.9%, respectively, p < 0.01). Delayed notifications were significantly higher for biochemical parameters (19.7%, p < 0.001) and observed particularly in morning hours (06:00 a.m.–10:00 a.m.), lunch break time (12:00–14:00) and end of the working day (16:00–18:00). Latencies of mild-delayed reporting were 18.5 ± 4.4 min for 62.8% and advanced-delayed reporting were 47.1 ± 11.3 min for 37.2% of total delayed notifications. Most of the critical results were reported to the health care staff other than physician (55.6%). Conclusion: Laboratory professionals should work in col-laboration with responsible clinician and healthcare staff in critical value reporting process.


Introduction
Critical or panic values were first described by Lundberg as "potentially life-threatening test results unless necessary treatment is done promptly in clinical laboratories" [1,2]. Reporting and documentation of critical values has been performed for patient safety in clinical laboratories and written procedures for reporting critical values are recommended as a quality requirement by national regulatory authorities. Many clinical laboratories have prepared and used their quality procedures for defining critical values, their limits, reporting time and documentation process [3][4][5]. Various reporting systems have gradually developed for effective notification including phone calls, online reporting and text messages, but all have some limitations [6,7]. Delays in critical value reporting due to these limitations are considered as laboratory error and may lead to increases in mortality and morbidity. Lack of access to healthcare staff can be an important limitation and lead to delays in reporting critical values. It also can cause overloading of laboratory staffs because of recurrent phone calls.
Despite its importance there are few studies in literature evaluating delaying times and causes of delays in critical value notification [8,9] and to our knowledge no study in Turkey.
In this study we aimed to investigate the frequency of delayed notifications and probable causes of delays in critical value notification respective to the time of the day and related clinical departments in our university hospital.

Materials and methods
All data were collected from reports generated by laboratory information system (LIS) and written documents during the 2 months period in biochemistry laboratory of Mustafa Kemal University, beginning in May 2015. Written consent was obtained from hospital administration.
During this observational study, we aimed to analyze delayed notifications. Therefore, all critical values were considered as separate events. Result generation time and reporting time for each critical value were recorded separately and delayed tests and delay times were evaluated daily. Time intervals exceeding 15 min between critical value alarming time and reporting time were considered as mild-delayed notification and exceeding 30 min were considered as advanced-delayed notification [8].
Results were analyzed in terms of test origin and frequency of critical value reporting and healthcare personnel to whom the result was notified by phone call. The rate of delayed notifications, delay times and working hours at which delaying notifications occurred were also evaluated. Regulatory-preventive action forms (RPAF) were prepared for delayed notifications in the end of the study as a requirement for routine quality procedure in our laboratory. Responsible service nurses and clinicians were questioned for the probable reasons of late answering.
Statistical analyses were performed using statistical software (SPSS for Windows, verison 21.0, Chicago, IL, USA). Critical value, delayed reporting frequencies and their percentages were calculated and statistical difference was determined by χ 2 -test. p < 0.05 was accepted as statistically significant.  (Figure 1). Most frequently delayed notifications were observed between 6:00 a.m and 8:00 a.m. and their distributions through working hours are shown in Figure 2. Delayed notifications were higher in biochemistry parameters (19.7%) compared to hematology (12.6%) and blood gas parameters (5.6%) (p < 0.001). Nearly more than half of the critical values (55.6%) were reported to non-physician personnel (21.6% for nurses, 17.6% for intern doctors and 16.1 for service secretaries and 44.4% of total critical values were reported to the responsible physicians in our hospital (Figure 3).

Discussion
In this study, we evaluated the delaying frequencies and delay times for critical value reporting by phone call in the clinical biochemistry laboratory. Reducing preventable laboratory errors is an important goal for the clinical laboratories [10]. This evaluation provides a context   for prevention of delayed notification and improvement of critical value reporting because effective reporting impacts on patient safety and efficiency of interaction between clinics and clinical laboratories. A total of 2018 critical values were reported by phone call and documented in our laboratory during the 2-month study. As shown in Table 1, critical value frequency was approximately two times higher in tests ordered from service and ICU patients than those of polyclinics similar to previously reported by Dighe et al. [8]. They found 3.5 times higher critical callbacks for inpatient tests than outpatients. This is an expected result because patients in services especially in ICU's have more life-threatening values than the outpatients. We also observed that the mean delaying times were 18.5 ± 4.4 min for 62.5% and 47.1 ± 11.3 min for 37.5% of total delayed tests. In literature there is no consensus about the exact time period for an appropriate reporting time. Some studies have recommended that critical values should be reported within 15-30 min after testing is completed [8,11]. Some others reported higher durations (30-120 min) [12]. After consulting responsible clinicians, we consider that these threshold limits seem appropriate for our hospital standards. 13.1% of total critical value notifications by phone call were accepted as delayed notifications and most of them for biochemistry and hematology parameters because of late answering (also called as dropped call) by responsible healthcare staff. Delayed reports from blood gas analyses were less frequent than biochemical and hematology parameters. Delayed notifications were remarkably higher in services and polyclinics patients compared to those of emergency departments and intensive care units (Table 1). It is well known that critical value reporting is an important issue because of patient safety and test results (e.g. blood gas parameters) of critically ill patients in these departments (ED and ICU) are monitored closely by the nursing staff and clinicians. Therefore, these lower percentages can be resulted from the nursing staff working in these departments being more aware of the importance of critical values compared to services and polyclinics. In literature, different dropped call rates between 0.1 and 0.3% have been reported previously [13]. Our rates were higher than those of reported previously. To find out the reasons of dropped calls, we analyzed the distribution of delayed notifications throughout the 24-h day (Figure 2). Most of the delayed notifications were prominent from 06:00 a.m. until 10:00 a.m., at the beginning of working hours which are the busiest times of a work-day. Other delayed notifications were observed between 12:00 and 14:00, lunch break time, and between 16:00 and 18:00, shift change time for nursing and other healthcare staff.
The most three prominent reasons we observed from RPAF's were preparation of patients for morning visits, beginning of phlebotomy process for test ordering and personnel task switching occurred during this period. We concluded that relatively increased workload at these hours and ineffective workflow planning during shift times can cause late answering for critical callbacks by healthcare staffs in our hospital.
Many laboratories have used different alerting (visual or audible) and reporting systems such as phone call, online reporting and text message for preventing delays in critical value reporting [7,14]. Nevertheless, phone call is still the most common method used in clinical laboratories because of its ease of use for asking all recipients to read back the critical values [15,16]. In a study Piva et al. recommended alternative reporting system ensuring read-back procedure. They used a software which generates automated notification by a specific text message sent to the cell phone of the responsible healthcare personnel. The alert message is appeared on the monitor until the responsible clinician acknowledges its receipt. If delay occurs within the timeliness, the notification is communicated by phone call [17]. Reporting of critical values through LIS to the smart phones of the responsible personnel of the clinics can be another alternative for critical value reporting. Because smart phones and its commercial applications (message groups or networks) have gradually increased. Additionally, some of them also include read-back applications. We have used phone call for critical value notification in our laboratory because lack of specific softwares mentioned above and "read back" is not possible by online reporting system used in our laboratory.
The other issue in critical value reporting relates to who should receive the phone call when laboratory technician reports out a critical value. In our hospital, critical values are reported to the intern doctors, nurses or service secretary by phone call if the responsible clinician is not available. In this study we found that 44.4% of critical values were reported to the responsible clinician. Remaining 55.6% of total notifications were made to the other healthcare staff. In a study, Howanitz et al. reported that nursing staff surveyed were not aware of critical values and did not think the calls for critical value were helpful. Whereas, majority of the clinicians surveyed were aware of the critical values and thought the calls were helpful [15]. However, the responsible clinician may not always be available because of increased workload or being in different locations (polyclinics, services or operating room) during working hours. Therefore critical value reporting to the non-physician nursing staff is inevitable in our hospital condition. We speculate that critical value education to increase awareness of health care staff about the importance of critical values in patient safety may be helpful in preventing delayed reports.
Critical value reporting is a medically important issue and its procedure varies between the laboratories and hospitals.
In conclusion, each laboratory should determine their critical value reporting policy according to their needs and facilities. However, there is a need for developing more effective technologies supporting phone call and decreasing workload of nursing staffs to maximize clinical benefits. Wireless technology usage (i.e. wireless phone) can be a useful choice because it is portable and can make clinicians always available. We also recommend that laboratory professionals should involve not only healthcare staffs but also administrative staffs who are responsible for personnel switching in developing critical value reporting policy to ensure patient safety.
Limitations, this study has focused on only delays between result generation time and reporting time for critical values. We did not evaluate the delays for overall testing process starting from sampling time. Further research is needed to evaluate this process more comprehensive.