Professional Documents
Culture Documents
Key Words: Informatics; Clinical pathology; Clinical chemistry; Management; Administration; Hematology; Patient safety; Critical value
reporting
DOI: 10.1309/R53XVC2U5CH6TNG8
only a few analytes for short periods or have reviewed a the chemistry, hematology, and outpatient health center lab-
small number of critical values in a number of different insti- oratories (chemistry and hematology) was included in our
tutions.5-7 In the present study, we analyzed 12 months of critical value analysis. Microbiology critical values were not
critical value data and more than 37,000 individual critical included in the present study because our microbiology lab-
results to understand the scope of critical value reporting and oratory uses a different documentation process for critical
identify opportunities for process improvement. values.
❚Table 1❚
Massachusetts General Hospital Critical Value List
Results ❚Table 2❚
Critical Values by Laboratory
❚Table 3❚
Critical Values by Test
2,500
ED
Outpatient
2,000 Inpatient
1,000
500
0
0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
9-10
10-11
11-12
12-13
13-14
14-15
15-16
16-17
17-18
18-19
19-20
20-21
21-22
22-23
23-24
Hour of Day
❚Figure 1❚ Critical values vs time of day. Distribution of critical value calls vs time (24-h clock) for the emergency department
(ED), outpatients, and inpatients.
rather may be communicated to the caregiver via an opera- mean, 0.5 minute; non-ICU settings mean, 2.0 minutes; P =
tions associate (OA; clerical staff members who perform .010). This was likely due to the greater availability of care-
clinical support functions). We therefore wanted to exam- givers in the ICU setting.
ine the communication process from the OA to the respon-
sible caregiver (physician or nurse). To examine the timeli- Critical Value Turnaround Time
ness of reporting, we created a logbook that each inpatient The “in-laboratory” turnaround time for each critical
floor maintained to monitor critical values. The OA docu- value was determined to assess the timeliness of critical value
ments in the logbook the time the call was received from reporting. For the 37,503 critical values, the mean time from
the laboratory, patient identifiers, the test result, and the the value entering the critical callback queue to the time when
time the critical result information was communicated to the critical value information was conveyed to the patient
the responsible caregiver. Critical results communicated location or ordering clinician was 22 minutes, and the median
directly to the licensed caregiver (without involving the time was 9 minutes (data not shown). Delays in critical value
OA) were not documented in the logbook because this reporting correlated with testing performed on outpatients and
information is already captured in the LIS callback appli- testing ordered on requisitions lacking the name of the order-
cation. We examined the logbooks for 29 inpatient care ing clinician or the ordering location. Tests performed in set-
units for 1 month. During this period, 1,477 critical values tings where there is continuous technologist presence (eg,
were documented in the logbook. The mean on-floor com- blood gases) were called back faster than tests performed in
munication time (OA to responsible caregiver) was 1.8 min- other areas. This information was useful as we began to imple-
utes (median, 1.0 minute). Increased communication times ment measures to improve critical value reporting in all areas
were observed on non-ICU floors compared with ICUs (ICU of the laboratory.
❚Table 4❚
Critical Values by Site
Location No. (%) of Critical Test Results No. of Beds Critical Values per Year per Bed
Examination of Critical Value Limits callback policies to reduce the number of critical callbacks.
To better understand our present upper and lower value For example, changing the lower limit callback value for glu-
limits for critical callbacks (eg, the limits for potassium of cose from less than 60 mg/dL (<3.3 mmol/L) to less than 45
<2.8 and >6.0 mEq/L [<2.8 and >6.0 mmol/L]), we plotted the mg/dL (<2.5 mmol/L) has resulted in 2,136 fewer calls per
number of critical callbacks for each analyte vs the result year (an overall reduction of 5.7% of all callbacks) ❚Figure 3❚.
value. This enabled us to examine the potential effect that
changing the limits of critical callback would have on call vol-
umes. Representative graphs of the critical value calls for
Discussion
potassium and glucose are shown in ❚Figure 2❚. This informa-
tion was used in conjunction with published literature and In this report, we provide a comprehensive view of the
consultation with clinicians to propose changes to the critical critical value reporting process in a large academic medical
A
120
100
No. of Critical Values
80
60
40
20
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
500-520
521-540
541-560
561-580
581-600
601-620
621-640
641-660
661-680
681-700
701-720
721-740
741-760
761-780
781-800
801-900
901-1,000
>1,000
Test Result Value
B
1,600
1,500
1,400
1,300
1,200
No. of Critical Values
1,100
1,000
900
800
700
600
500
400
300
200
100
0
0.2
0.5
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
7.0
7.1
7.2
7.3
7.4
7.5
7.6
7.7
7.8
7.9
8.0
8.1
8.2
8.3
8.4
8.5
8.6
8.7
8.8
8.9
9.0
9.1
9.2
9.3
9.4
9.5
9.6
9.7
9.8
9.9
10
❚Figure 2❚ Critical value limit analysis for glucose and potassium. The number of critical value callbacks for glucose (A) and
potassium (B) are plotted vs the test result value. The dotted line indicates the critical value limits for each test (glucose, <45 or
>500 mg/dL [<2.5 or >27.8 mmol/L]; potassium, <2.8 or >6.0 mEq/L [<2.8 or >6.0 mmol/L]).
center. We provide details regarding the scope, volume, tim- list. In addition to determining which tests are to be included
ing, and operational aspects of critical value reporting. Many in the critical values list, another important strategy is to
of these parameters should be applicable to a variety of set- examine the consequences of changing the boundaries for crit-
tings. This analysis provides a context for comparison and ical value reporting. These boundaries must be defined in con-
process improvement. sultation with clinicians. Small changes in critical value
Increasing workload in the clinical laboratory makes it reporting parameters may result in the addition or loss of thou-
important to achieve efficient use of laboratory resources to sands of phone calls for the laboratory staff.
maximize clinical benefits. Expansion of critical callback lists Outpatient critical values present unique challenges in
to include testing that does not meet the criterion of the timely reporting to clinicians. One of the strongest correlates
“imminent danger” standard may dilute the urgency of a crit- of delayed reporting of critical values was the specimen being
ical value call and lead to unnecessary interruptions for clini- obtained from an outpatient. Outpatient critical values are
cians. For example, critical value calls for high creatinine lev- challenging to communicate to the responsible clinician
els will not be of clinical value for patients receiving dialysis because there often are different approaches in various prac-
and in many situations in which the high creatinine value is an tices for determining patient coverage. Unlike inpatients, there
expected finding. In addition, there are many clinical settings is no fixed patient location that can be phoned.
(chemotherapy, malignancy) in which the “critical” test result Another factor we identified as causing delays for outpa-
is expected and reporting of this value may not contribute to tients was illegible or missing ordering provider information. As
improved patient care. By applying this logic to other scenar- a result of this analysis, we have changed our medical policy to
ios, we have not adopted critical callbacks for positive cardiac explicitly state that all requisitions must have an ordering
markers (creatine kinase-MB and troponin T). The marginal provider and an ordering location printed on the requisition.
clinical usefulness vs the marginal resource cost should be We are in the process of communicating this to our caregivers.
considered carefully when the tests and cutoff limits for criti- We also have instituted daily exception reports of critical val-
cal value reporting are determined. National standards have ues called back in times that exceed our threshold limit of
been published concerning critical value ranges.8,9 These stan- acceptability (30 minutes). These reports are distributed to the
dards provide a benchmark against which the laboratory can laboratories and are being used to understand and remedy the
compare and adjust its critical values list accordingly. root causes of delays in critical results reporting. We have noted
Communication by telephone, especially when per- that recent improvements in the critical value communication
formed by technologists, is a costly practice in terms of the times have coincided with increased awareness of critical value
resources required to complete the phone calls and document monitoring. We presently are working with our outpatient
the process. For this reason, it is helpful to try to reduce the practices to improve communication between the laboratories
number of phone calls by careful review of the critical values and the outpatient care centers.
1,000
800
No. of Results
600
400
200
0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Glucose Value (mg/dL)
❚Figure 3❚ Glucose test results and critical value limits. Glucose values are plotted vs the number of results with that value for a
12-month period. The dotted lines indicate the 2,136 results falling between the previous critical value limit (<60 mg/dL [<3.3
mmol/L]) and the current critical value limit (<45 mg/dL [<2.5 mmol/L]). To convert conventional values (mg/dL) to Système
International units (mmol/L), multiply by 0.0551.
Another contributor to delays in outpatient critical value because many of these “critically low” results will be falsely
reporting is the heterogeneity of the outpatient population, low or no longer relevant. The ability to provide a physician-
with specimens arriving from health centers, clinics, urgent specific critical values list could eliminate a large number of
care centers, dialysis centers, and physicians’ offices. Each of unnecessary critical value calls. These systems, when inter-
these areas is likely to have a different call schedule, answer- faced with automated alerting systems, will have the potential
ing service, and cross-coverage procedure, making reliable to improve patient safety and provide more context-sensitive
communication with the responsible licensed caregiver chal- critical value reporting. At present, practical implementation
lenging. The nature of outpatient specimen transport and pro- of this scenario would be constrained by regulations (partic-
cessing often results in outpatient test results being generated ularly the JCAHO National Patient Safety Goals) that require
in the evening when the outpatient clinic or physician’s office all critical results to be communicated and do not allow for
is closed. The laboratory must have a mechanism to determine more nuanced approaches.
on-call coverage and work with outpatient practices to
improve the communication processes. From the Departments of 1Pathology, 2Medicine, and 3Nursing,
The potential for technological solutions to improve the Massachusetts General Hospital, Boston.
process of critical value reporting is evident in numerous Address reprint requests to Dr Lewandrowski: Clinical
reports.10,11 The use of information technology to automatical- Chemistry, Bigelow Bldg, Room 510, Massachusetts General
ly communicate with the responsible provider has been Hospital, 55 Fruit St, Boston, MA 02114.
demonstrated to reduce the critical value reporting time in
controlled settings. For implementation of automated critical
value reporting, interfaces from the LIS to technologies that References
facilitate bidirectional communication (such as e-mail or 2-
1. Lundberg GD. When to panic over abnormal values. MLO
way pagers) need to be developed. An important component Med Lab Obs. 1972;4:47-54.
in such a system is the ability of the automatic reporting sys- 2. Lundberg GD. Critical (panic) value notification: an
tem to reliably determine the identity of the responsible established laboratory practice policy (parameter) [editorial].
provider. At larger medical centers, this task can be challeng- JAMA. 1990;263:709.
ing because there may be different coverage lists, tests ordered 3. Joint Commission on the Accreditation of Healthcare
Organizations: National Patient Safety Goals. Available at:
by consultants unknown to the primary caregiver, and patient http://www.jcaho.org/accredited+organizations/patient+safety/
transfers to different locations. An electronic reporting system npsg.htm. Accessed August 22, 2005.
potentially could create dangerous delays in communication if 4. Center for Medicare and Medicaid Services, Department of
not properly implemented. The system needs to have an health and Human Services. Clinical Laboratory
Improvement Amendments of 1988. 68 Federal Register 1047
“acknowledgment” function such that the laboratory can (2003) (codified at 42 CFR §493.1291(g)).
ensure that the responsible caregiver received the result. 5. Lum G. Assessment of a critical limit protocol for point-of-
Electronic systems also require an escalation procedure so that care glucose testing. Am J Clin Pathol. 1996;106:390-395.
lack of acknowledgment of the critical result prompts an alter- 6. Howanitz PJ, Steindel SJ, Heard NV. Laboratory critical values
native approach for communication. policies and procedures: a College of American Pathologists
Q-Probes study in 623 institutions. Arch Pathol Lab Med.
Development of LIS middleware with alert reporting 2002;126:663-669.
software should permit highly nuanced approaches to critical 7. Kuperman GJ, Boyle D, Jha A, et al. How promptly are
value reporting in the near future. Rules-based logic can be inpatients treated for critical laboratory results? J Am Med
applied to laboratory values to build alerts that take into Inform Assoc. 1998;5:112-119.
account not only the result value, but also other related 8. Kost GJ. Table of critical limits. MLO Med Lab Obs.
2004;36:6-7.
results, a change in the current test result from previous
9. Burtis CA, Ashwood ER, Bruns DE. Tietz Textbook of Clinical
results (ie, delta checks), patient demographics, ordering Chemistry and Molecular Diagnosis. Philadelphia, PA: Saunders;
provider, and other parameters to customize the alerting to 2005.
the patient’s condition and the needs of the clinical team for 10. Bates DW, Pappius E, Kuperman GJ, et al. Using information
notification. For example, many oncology physicians do not systems to measure and improve quality. Int J Med Inform.
1999;53:115-124.
want to be notified regarding patients with neutropenia.
11. Tate KE, Gardner RM, Weaver LK. A computerized laboratory
Similarly, there is little usefulness in notifying a diabetologist alerting system. MD Comput. 1990;7:296-301.
of low glucose values for patients seen in an outpatient clinic