SEARCH

SEARCH BY CITATION

Keywords:

  • emergency department;
  • laboratory turnaround time;
  • emergency department crowding

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References

Objectives:  Increases in emergency department (ED) visits may place a substantial burden on both the ED and hospital-based laboratories. Studies have identified laboratory turnaround time (TAT) as a barrier to patient process times and lengths of stay. Prolonged laboratory study results may also result in delayed recognition of critically ill patients and initiation of appropriate therapies. The objective of this study was to determine how ED patient volume itself is associated with laboratory TAT.

Methods:  This was a retrospective cohort review of patients at five academic, tertiary care EDs in the United States. Data were collected on all adult patients seen in each ED with troponin laboratory testing during the months of January, April, July, and October 2007. Primary predictor variables were two ED patient volume measures at the time the troponin test was ordered: 1) number of all patients in the ED/number of beds (occupancy) and 2) number of admitted patients waiting for beds/beds (boarder occupancy). The outcome variable was troponin turnaround time (TTAT). Adjusted covariates included patient characteristics, triage severity, season (month of the laboratory test), and site. Multivariable adjusted quantile regression was carried out to assess the association of ED volume measures with TTAT.

Results:  At total of 9,492 troponin tests were reviewed. Median TTAT for this cohort was 107 minutes (interquartile range [IQR] = 73–148 minutes). Median occupancy for this cohort was 1.05 patients (IQR = 0.78–1.38 patients) and median boarder occupancy was 0.21 (IQR = 0.11–0.32). Adjusted quantile regression demonstrated a significant association between increased ED patient volume and longer times to TTAT. For every 100% increase in census, or number of boarders over the number of ED beds, respectively, there was a 12 (95% confidence interval [CI] = 9 to 14) or 33 (95% CI = 24 to 42)-minute increase in TTAT.

Conclusions:  Increased ED patient volume is associated with longer hospital laboratory processing times. Prolonged laboratory TAT may delay recognition of conditions in the acutely ill, potentially affecting clinician decision-making and the initiation of timely treatment. Use of laboratory TAT as a patient throughput measure and the study of factors associated with its prolonging should be further investigated.

ACADEMIC EMERGENCY MEDICINE 2010; 17:501–507 © 2010 by the Society for Academic Emergency Medicine

The number of patients presenting for emergency department (ED) care has been rapidly increasing. According to the National Hospital Ambulatory Medical Care Survey, there were 119.2 million ED visits in 2006, up from 90.3 million (32% increase) just 10 years earlier.1 A consequence of increasing numbers of patients and decreased number of facilities capable of treating these patients is crowding in the nation’s EDs.2,3 ED crowding is associated with poorer patient satisfaction, lower quality of care, and both treatment delays and nontreatment of acute painful conditions.4–10 In addition, ambulance diversion secondary to crowding is associated with an increased door-to-needle time for patients receiving thrombolytic therapy for acute myocardial infarction,4 while prolonged ED boarding has been associated with a higher risk of death in critically ill patients.11

Increases in ED visits may place a substantial burden on hospital-based laboratories. Studies have identified laboratory turnaround time (TAT) as a barrier to patient process times and increased lengths of stay.12,13 Conversely, it is possible that critical laboratory results may be delayed during episodes of ED crowding. Such delays can be potentially hazardous because critically ill patients may go unrecognized and, as a result, experience treatment delays and worse outcomes.

To study the relationship between ED crowding and hospital laboratory TAT, we evaluated the association between ED patient volume and troponin laboratory TAT (TTAT). TTAT is an ideal process of care to evaluate because the timely availability of these results guide clinicians in patient care decision-making, initiation of hospital admissions, and treatment for conditions such as acute coronary syndromes.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References

Study Design

This was a multicenter, retrospective cohort review. This study was approved by the institutional review boards of the participating sites.

Study Setting and Population

The study setting and population consisted of ED visits at five academic, tertiary care hospitals located in the Northwest, West, Northeast, and Mid-Atlantic regions of the United States. These study sites were selected for participation to provide geographic and ED (i.e., trauma level) diversity. All participating institutions have central laboratories where both ED and hospital specimens are processed. All EDs have electronic health records (EHRs). To capture seasonal variation, data were collected on all adult patients seen in each ED with cardiac troponin testing during the months of January, April, July, and October 2007.

Study Protocol

Data were collected using administrative reports created by the EHRs utilized in these EDs. All five EDs have comprehensive EHRs for patient tracking, physician and nursing documentation, order entry, and reporting of test results. All data entered into the systems, including when troponin tests are ordered and resulted, are electronically time stamped. Negative (i.e., result times preceding the ordering times) or aberrant TTAT values (i.e., result times after the patient was discharged, and >24 hours in duration) were considered bad data and thus excluded from the analysis.

Primary predictor variables were ED crowding factors. Because there is no criterion standard metric for ED crowding,14 a priori two measures of ED patient volume were used: 1) number of all patients in the ED divided by the number of licensed beds (occupancy) and 2) number of admitted patients waiting for beds divided by number of licensed beds (boarder occupancy). Number of patients in the ED included patients in waiting rooms, and licensed beds included hallway beds and ED-based observation units. For those EDs with separate pediatric and psychiatric units and no crossover of clinical staff or resources, these types of patients were excluded in the patient volume counts. All predictor values were determined at the time the troponin test was ordered by the treating clinician.

Covariates for adjusted analyses included age, sex, race/ethnicity, patient triage severity index, site, and month of the patient visit for which the troponin test was ordered. During the study period, the Emergency Severity Index (ESI, a validated five-point acuity scale, 1 = acute, 5 = nonacute),15 was used as the triage severity index at four of the sites; one site used a four-level triage system (1 = most urgent, 4 = least urgent). A comparison of the five-level ESI distributions of the four sites with that of the single four-level system revealed similar distributions at the nonacute and least urgent levels. For uniformity of the triage severity levels, Levels 4 and 5 of those sites using the five-level ESI were collapsed to Level 4. Any missing covariate data (e.g., race/ethnicity, triage severity) were coded as “missing” to assess whether “missingness” was associated with the outcome (there was no association).

Key Outcome Measure

The outcome variable of interest was laboratory TTAT. This was defined as the minute time from when the troponin test was ordered to when the results were available and posted to the ED’s EHRs.

Data Analysis

Descriptive statistics by site are presented for cohort characteristics, ED crowding indicators, and TTATs. Multiple models were evaluated using ED crowding indicators as dichotomous, categorical, and continuous values. TTAT outcome data are presented as continuous minute-time values, as these produced models with best fit. Covariates were determined a priori for construct validity (i.e., patient characteristics, triage severity, month of testing, site).

Because TTAT time data were severely right skewed and heteroscedastic, a log transformation would have been required to run regression models using either ordinary least squares with site treated as a fixed effect or hierarchical linear models with site treated as a random effect. To avoid the potential bias that could be introduced when transforming the results back to the unlogged scale even when applying a smear factor in the presence of heteroscedasticity,16 quantile regression was used.17 In the same way that linear regression models the conditional mean of the response variable, quantile regression can be used to model specific percentiles of the response distribution conditional on the predictors, including the median (the 50th percentile). The median is a special quantile that is often used as a measure of central tendency, especially when the data are skewed, because unlike the mean, it is not sensitive to outliers. Median regression (i.e., quantile regression of the 50th percentile) is therefore suitable for time data that are frequently nonnormal in distribution.17,18 Unlike linear regression that can only model the conditional mean, quantile regression can be used to model noncentral locations of the conditional response, such as the lower (e.g., 10th percentile) and upper (90th percentile) ends of the response variable distribution, allowing the researcher to assess whether the effect of a predictor remains constant or changes (e.g., whether the effect of crowding is the same for short TTATs as long TTATs). For this study, in addition to the median (50th percentile), the 10th, 25th, 75th, and 90th percentiles were run. The quantile regression coefficient estimates the change in the modeled quantile for every one-unit change in the predictor variable, so if the median is modeled, the coefficient is the change in median for every one-unit change in the independent variable. All analyses were completed using SAS (Version 9.1, SAS Institute, Cary, NC). Quantile regression was carried out with the SAS QUANTREG procedure, using the simplex algorithm and bootstrapped standard errors and confidence limits. Results presented are those of models with best fit.

Additional time data were available for the time the specimen was received by the hospital laboratory at one participating study site (Site 1). Because TTAT is defined as the time from when the troponin test was ordered to when the result was available, the time the specimen was received by the laboratory allows for the calculation of two additional time outcomes, dividing the TTAT into: 1) the time from when the troponin test was ordered to when the specimen was received by the laboratory (ED collection time) and 2) the time when the specimen was received to when the result was available (lab processing time; i.e., ED collection time + lab processing time = TTAT). Site 1 subanalyses were completed of ED collection time and lab processing time outcomes using the methods, ED crowding predictors, and covariates described above.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References

During the study period, there were a total of 99,279 ED visits at the five study sites. Of these, 9,492 (10%) had routine troponin tests ordered in the ED. Across five sites, patient cohort characteristics included mean ages 57–64 years; 50% to 54% female patients; 27% to 77% white patients, 7% to 65% Black or African American patients, 0% to 29% Hispanic patients; and the median ESI (modified to four levels) was 2 (interquartile range [IQR] = 2–3). For the outcome of TTAT, 272 (3%) specimens had negative time values, and 25 (0.3%) had result times posted that exceeded 24 hours and were after the patient was discharged from the ED. Mean TTAT values (excluding these times) at each site ranged from 69–149 minutes, with a cohort median of 107 minutes (IQR = 73–148). For primary ED crowding predictor variables, the median occupancy for this cohort was 1.05 patients (IQR = 0.78–1.38), and median boarder occupancy was 0.21 (IQR = 0.11–0.32). Characteristics of the participating sites and entire study cohort are provided in Table 1.

Table 1.    Characteristics of Study Cohort and Participating Sites
 Overall Study Cohort (N = 9,492)Site 1 (n = 2,766)Site 2 (n = 1,335)Site 3 (n = 3,048)Site 4 (n = 1,838)Site 5 (n = 505)
  1. IQR = interquartile range; TAT = turnaround time (time ordered to time resulted); TTAT = troponin turnaround time.

  2. *Results adjusted for aberrant troponin TAT values (results times posted after ED discharge times and >24 hours duration AND negative values excluded).

Age, yr
 Mean (SD)60 (17)64 (17)58 (17)58 (17)57 (16)59 (16)
 Median (IQR)59 (48–72)64 (51–78)57 (46–70)56 (46–70)55 (46–67)58 (47–72)
Female, n (%)4,992 (53)1,506 (54)664 (50)1,633 (54)932 (51)257 (51)
Race/ethnicity, n (%)
 White3,233 (34)745 (27)1,022 (77)652 (21)526 (29)288 (57)
 Black or African4,134 (44)796 (29)99 (7)1,967 (65)1,146 (62)126 (25)
 American1,026 (11)807 (29)66 (5)32 (1)50 (0)71 (14)
 Hispanic162 (2)39 (1)54 (4)46 (2)17 (0)6 (1)
 Asian533 (6)93 (3)10 (1)320 (10)99 (0.1)11 (2)
 Other404 (4)286 (10)84 (6)31 (1)3 (0.1)
 Unknown/missing      
ESI or triage level, n (%)
 1410 (4)48 (2)11 (1)318 (10)21 (1)12 (2)
 25,682 (60)1,797 (65)418 (32)2,287 (75)950 (52)230 (46)
 33,207 (34)888 (32)850 (65)419 (14)798 (43)252 (50)
 4145 (2)32 (1)22 (2)20 (1)61 (3)10 (2)
 513 (0)1 (0)3 (0.2)8 (0.4)1 (0)
 Missing35 (0.4)31 (2)4 (0.1)
TTAT (min)*
 Mean (SD)121 (82)149 (99)69 (55)132 (71)112 (65)80 (68)
Median (IQR)107 (73–148)132 (101–168)58 (44–78)121 (92–158)94 (70–129)66 (51–92)
Trauma center levelLevel 2Level 1Level 1Level 1Level 2
 Locale UrbanSub/urbanUrbanUrbanSub/urban
Annual visit volume in 200776,00040,00056,00061,00055,000
ED square footage9,40010,10228,00012,50014,747
Number of adult licensed ED beds4239373742
Census
 Mean (SD)44 (20)62 (22)27 (8)40 (12)38 (13)27 (13)
 Median (IQR)41 (29–53)61 (46–79)27 (21–32)41 (32–49)39 (29–48)26 (17–37)
Number of boarders
 Mean (SD)9 (7)14 (7)8 (3)4 (3)12 (5)5 (4)
 Median (IQR)8 (4–13)13 (9–19)8 (5–10)4 (2–6)12 (8–16)4 (1–8)
Occupancy
 Mean (SD)1.10 (0.48)1.45 (0.52)0.69 (0.20)1.09 (0.32)1.04 (0.35)0.61 (0.30)
 Median (IQR)1.05 (0.78–1.38)1.40 (1.05–1.83)0.69 (0.54–0.82)1.11 (0.86–1.32)1.04 (0.78–1.30)0.57 (0.38–0.83)
Boarder occupancy
 Mean (SD)0.23 (0.17)0.34 (0.17)0.20 (0.09)0.12 (0.08)0.34 (0.14)0.11 (0.10)
 Median (IQR)0.21 (0.11–0.32)0.31 (0.21–0.45)0.21 (0.13–0.26)0.11 (0.05–0.16)0.32 (0.22–0.43)0.07 (0.02–0.17)

The adjusted quantile regression results (Table 2) indicate with this cohort, for every unit increase of our standardized metric occupancy (i.e., a 100% increase of census over the number of beds in each ED), there was a 12 (95% CI = 9 to 14)-minute increase in the median (50th percentile) TTAT. For every unit increase in boarder occupancy (e.g., a 100% increase in number of boarders in each ED beyond bed capacity), there was a 33 (95% CI = 24 to 42)-minute increase in the median (50th percentile) TTAT. These associations were greater at the 90th percentile of TTAT, whereby for every unit increase in census/beds or number of boarders/beds there was an increase in TTAT of 15 (95% CI = 6 to 25) minutes or 62 (95% CI = 34 to 91) minutes, respectively. Patient related covariates of age, sex, race/ethnicity, and month were not significant in any of the adjusted multivariable analyses.

Table 2.    Multivariate Adjusted Quantile Regression Analyses of TTAT*
  1. *Adjusted for age, sex, race, Emergency Severity Index score, month, and site.

  2. TTAT = troponin turnaround time; occupancy = census/number of beds; boarder occupancy = number of boarders/number of beds.

ED Crowding Patient Volume Predictors (per Unit Increase)TTAT Parameter Estimates, minutes (95% CI)p-value
Occupancy
 10th percentile of TTAT8.88 (6.47–11.29)<0.0001
 25th percentile of TTAT12.12 (9.72–14.51)<0.0001
 50th percentile of TTAT11.54 (8.92–14.17)<0.0001
 75th percentile of TTAT8.35 (4.74–11.95)<0.0001
 90th percentile of TTAT15.42 (6.07–24.76)0.001
Boarder occupancy
 10th percentile of TTAT23.90 (17.43–30.38)<0.0001
 25th percentile of TTAT31.24 (25.09–37.40)<0.0001
 50th percentile of TTAT32.67 (23.83–41.50)<0.0001
 75th percentile of TTAT33.42 (19.82–47.01)<0.0001
 90th percentile of TTAT62.44 (33.91–90.98)<0.0001

Using quantile regression analyses for Site 1 data, where laboratory receipt times of the troponin specimens were available, a comparison of ED collection time and laboratory processing time (serial subcomponents of TTAT) found significantly longer processing times in association with both occupancy and boarder occupancy. For every unit increased in our standardized metric census/beds (i.e., for every 100% increase in census or an additional 42 patients in this 42-bed ED), median ED collection time increased by 4 (95% CI = 2 to 6) minutes and median laboratory processing time increased by 8 (95% CI = 5 to 12) minutes. For every increase in number of boarders/beds (i.e., for every 100% increase in number of boarders in this ED or an addition of 42 boarders in this 42-bed ED), median ED collection time increased by 14 (95% CI = 7 to 21) minutes, and laboratory processing time increased by 25 (95% CI = 13 to 36) minutes.

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References

In this multicenter study, ED crowding was associated with longer TTATs. Regardless of how ED crowding was measured (whether as a total number of ED patients divided by number of ED beds or number of admitted ED patients waiting for hospital beds divided by number of ED beds, census/licensed ED beds, or boarders/licensed ED beds), increased ED patient volume was significantly associated with longer TTATs. The implications of the study findings presented here are twofold. The first is the potentially negative effect delayed troponin tests may have on patient care, and the second is the association of ED crowding with slower ED throughput measures of efficiency.

Troponin cardiac biomarkers are a diagnostic test used in the detection of cardiac-specific cell injury in patients suspected of having cardiac conditions.19 Over a third of acute myocardial infarction patients, however, do not present with chest pain.20 For this reason, timely troponin results are often relied on for the initiation of appropriate therapies for patients with atypical chest pain. Implications of this study include the potential downstream effects of prolonged diagnostic test results for patients with unclear but acute cardiac conditions. The timely receipt of a troponin result is critical for the recognition of ischemic events. It may affect treatment and decisions about hospital admission, and if the patient is admitted, the type of hospital bed he or she may require (nonmonitored vs. telemetry vs. critical care unit beds). It has been projected that efficient testing of TTATs could “facilitate earlier decision making, leading to more rapid disposition and treatment decisions … and thus have a salutary impact on options for interventions in patients with ACS or non-ST-segment elevation acute myocardial infarction.”21 Conversely, barriers such as slow laboratory turnaround in association with high ED patient volume may impede rapid assessment, recognition, and initiation of treatment for appropriate patients.22,23

From the perspective of studying delays in patient care, laboratory process time has been identified as an important solution to improve efficiency of patient flow and treatment within the ED.24 The Institute of Medicine report, Hospital-Based Emergency Care: At the Breaking Point, along with several other studies, has addressed laboratory testing of blood as a potential process improvement target to address operational efficiency.21,24,25 Simulation models have found that decreased laboratory TATs are associated with decreased ambulance diversion and length of stay and increased patient throughput.26 Recent studies indicate reduction of throughput process times may help to reduce ED lengths of stay,13,27,28 while other studies have even proposed use of ED ancillary service TAT as a potential measure for crowding itself.29 The results of this study demonstrate a direct association between ED occupancy and boarder occupancy with laboratory TTATs. From these results, it is possible that as ED volume increases, the number of troponin tests also increases under conditions of fixed resources (e.g., clinical and laboratory staffing, diagnostic equipment to run the laboratory specimens) such that demand exceeds supply. This in turn leads to a backlog of troponin specimens on which the test is to be run and the associated increase in TTAT. While speculative in nature, this proposed sequence of events is highly plausible because predictor values (occupancy and boarder occupancy) were calculated at the time a test was ordered (i.e., before the specimen was processed) as opposed to when it was resulted.

In an effort to address hospital ED throughput issues and develop quality measures and legislation designed to relieve crowding, the Federal Government Affairs Committee of the American College of Emergency Physicians (ACEP) has proposed recommendations to the ACEP Council.30 These include resolutions that ACEP advocate “emergency physicians be held accountable for those aspects of ED throughput time over which they have direct control.” Median time to troponin results (time from initial troponin order to result) is one of several measures proposed as an ED throughput performance measure,30 highlighting the importance of this metric whether linked to outcomes or efficiency. Results of this study indicate that longer TTATs are directly associated with increased ED patient volume—an aspect over which emergency physicians have no direct control. Further research is needed to better understand how factors intrinsic to either the ED or the hospital may influence proposed throughput measures such as TTATs. Conversely, if such a metric is implemented, system improvements to address increased ED patient volumes should be employed by hospitals to assist emergency physicians in meeting performance measures.

Emergency department crowding may be mistakenly regarded as a problem contained within the ED of an individual hospital or medical center. The prolonged TTATs of this study may have been entirely due to longer ED collection times. Subanalysis of Site 1, however, indicates that prolonged TTATs occurred both with ED collection and laboratory processing times. If the census or number of boarders in its ED exceeded bed capacity by 100%, median ED collection times increased by upwards of 4 and 14 minutes, respectively. The laboratory took even longer to process specimens, with an increase of 8 and 25 minutes in median times with respect to increased ED patients/beds or numbers of boarders/beds. While these individual site results are not generalizable, the limited findings here demonstrate that delays in laboratory TTATs were not restricted to the ED for this particular institution, but also occurred with the central laboratory. Based on this information, it is also possible that the increased TTAT processing times of the Site 1 laboratory were experienced not only by ED specimens, but also by specimens from other parts of the medical center that share the central laboratory facility. This could have potentially affected critical laboratory values for hospital patients, and timely clinical decision-making by other medical services that share the central laboratory. Unfortunately, it is not known whether longer processing times were the result of inefficiencies in the hospital, the ED, the laboratory, or all of the above. For Site 1, however, periods of crowding were associated with longer process times in both the ED and the hospital laboratory.

For this cohort, variation may have existed with how laboratory specimens are routinely handled at each site by the clinicians within the ED and then by the laboratory (i.e., differences in demand/capacity matching). It is not known how the aggregate, prolonged TTATs reported in this study may have translated into increased or decreased delays for individual patients. More importantly, however, these data demonstrate that while there is variability in crowding effects, its impact continued to be seen even at the lower quartiles of TTAT and generally with increasing magnitude at higher quartiles of TTAT (Table 2). There were differences in the magnitude of effect patient volume had on TTAT for the cohort. For example, for each 100% increase of census or number of boarders over bed capacity, there were large differences in effect when comparing the 10th percentile and the 90th percentile of TTAT (9 minutes vs. 15 minutes, respectively, for occupancy; 24 minutes vs. 62 minutes, respectively for boarder occupancy). Disparities of this magnitude in overall cohort processing times may have affected individual clinical decision-making.

For the purposes of this study, it must be emphasized that laboratory process time was used to gauge the association of patient care times with crowding in EDs. The speed with which the laboratory test was resulted (whether as the sum of collection time and laboratory processing time or separately for one site) was used as a process of care time outcome, not a quality of care measure. With regard to the selection of TTATs as the outcome of choice, the authors acknowledge that while these test results are critical for treatment and decision-making, it is not known what the direct significance is of longer laboratory TATs on patient-centered outcomes. Specifically for TTATs, it is not known what the consequences are of 10-minute, 30-minute, or 1-hour TATs. Guidelines currently do not exist delineating time frames for when a troponin test should optimally be resulted in association with improved patient outcomes. Whether TTAT may be considered time-sensitive, however, was not the objective of this study and should be further investigated.

Limitations

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References

Limitations to this study include the inability to determine how site-specific ED and hospital factors may have affected laboratory TATs. Data concerning ED-, hospital-, and laboratory-related factors (e.g., hospital inpatient census, ED staffing levels, number of other specimens being processed in the laboratory, variation with regards to laboratory equipment delays) specific to the time the troponin test was ordered were not available. While it is possible that increased laboratory TATs were the result of increased numbers of specimens subsequently being processed by the laboratory, it is not known how inpatient service delays may have also affected laboratory process times. The busyness of hospital services may have increased patient process times in the ED and as a result increased the number of patients in the ED waiting for test results.

This study did not investigate causal effects of the associations found. Time of day and day of week may have also affected TTAT processing time. While these data were available for four of the study sites (the institutional review board of one institution prohibited the sharing of detailed date-time data), we were unable to include these covariates in the analysis of the entire cohort. The statistical significance of the association between ED patient volume measures and TTAT did not change, however, in subanalyses including day of week and time of day for sites with available data.

While relevant patient-related characteristics such as age, sex, race/ethnicity, and ESI were used in adjusted analyses, information concerning chief complaints and comorbidities was not available. These types of patient-related characteristics, however, should not affect the outcome of choice here, TTAT. Patient factors such as chief complaints and comorbidity are more likely to influence physician decision-making to order a test and how quickly it is ordered after ED arrival. It is for this reason the outcome studied here was not the time from ED arrival to when a test was ordered or resulted, but instead the time when the test was ordered to when the result was available.

As with all retrospective studies, the documentation of patient care may not reflect actual times when activities were completed. Because all orders and results are time stamped in the EHRs of participating sites, however, the process times should be accurate. Additionally, “stat” troponins with point-of-care (POC) testing were not included in the analyses as these were available at only two of the five study sites. However, because POC diagnostic testing utilizes a different process (ED bedside testing by the treating clinician) than routine TTATs, which are collected and sent for hospital laboratory processing, the investigators also felt that it was appropriate to exclude this type of test from analyses. It is possible that POC testing may have affected the utilization of routine troponin laboratory ordering. For example, with high clinical suspicion of an acute condition or during periods of increased busyness, clinical staff may have preferentially used POC testing instead of the routine laboratory testing that takes longer to result. Such a scenario would make the results presented here conservative when compared to actual longer TATs.

The isolated findings of Site 1 ED collection and laboratory processing times may not generalize to other sites or settings. Future studies could include comparisons of other ED and hospital process times for laboratory or other diagnostic tests with sequential processes (e.g., radiographic studies, ED order time to arrival in radiology, and arrival in radiology to radiographic interpretation of results being available).

Finally, because this was a retrospective study, it was impossible to ascertain the actual TTAT of prolonged or negative TTAT data. Fortunately, because all data were generated from electronic medical record systems with time stamp values, the vast majority did not have these bad data values. A review of the small portion (<4%) of data with negative or prolonged values revealed these were isolated to two hospitals and over the span of 4 days, likely indicating a systematic error. A comparison of models run with and without the bad TTAT data also showed improving model fit when bad data were excluded.

Conclusions

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References

Crowding in EDs (as measured by patient volume) is associated with longer troponin laboratory turnaround times. Prolonged laboratory turnaround times such as these may delay recognition of conditions in the acutely ill and potentially affect clinician decision-making and initiation of timely treatment. Use of laboratory turnaround time as a patient throughput measure and the study of factors associated with its prolonging should be further investigated.

The authors thank Laura Rivera for assistance with data management.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References