American Journal of Emergency Medicine xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem

Original Contributions

Impact of case volume on outcome and performance of targeted temperature management in out-of-hospital cardiac arrest survivors☆ Seung Joon Lee, MD a, Kyung Woon Jeung, MD, PhD b, Byung Kook Lee, MD, PhD b,⁎, Yong Il Min, MD, PhD b, Kyu Nam Park, MD, PhD c, Gil Joon Suh, MD, PhD d, Kyung Su Kim, MD d, Gu Hyun Kang, MD e, for the Korean Hypothermia Network (KorHN) Investigators a

Department of Emergency Medicine, Myongji Hospital, Goyang, Gyeonggi-do, Republic of Korea Department of Emergency Medicine, Chonnam National University Hospital, Gwangju, Republic of Korea c Department of Emergency Medicine, School of Medicine, The Catholic University of Korea, Seoul, Republic of Korea d Department of Emergency Medicine, College of Medicine, Seoul National University, Seoul, Republic of Korea e Department of Emergency Medicine, Kangnam Sacred Heart Hospital, Hallym University, Seoul, Republic of Korea b

a r t i c l e

i n f o

Article history: Received 5 July 2014 Received in revised form 1 October 2014 Accepted 2 October 2014 Available online xxxx

a b s t r a c t Purpose: This study aimed to determine the effect of case volume on targeted temperature management (TTM) performance, incidence of adverse events, and neurologic outcome in comatose out-of-hospital cardiac arrest (OHCA) survivors treated with TTM. Methods: We used a Web-based, multicenter registry (Korean Hypothermia Network registry), to which 24 hospitals throughout the Republic of Korea participated to study adult (≥18 years) comatose out-of-hospital cardiac arrest patients treated with TTM between 2007 and 2012. The primary outcome was neurologic outcome at hospital discharge. The secondary outcomes were inhospital mortality, TTM performance, and adverse events. We extracted propensity-matched cohorts to control for bias. Multivariate logistic regression analysis was performed to assess independent risk factors for neurologic outcome. Results: A total of 901 patients were included in this study; 544 (60.4%) survived to hospital discharge, and 248 (27.5%) were discharged with good neurologic outcome. The high-volume hospitals initiated TTM significantly earlier and had lower rates of hyperglycemia, bleeding, hypotension, and rebound hyperthermia. However, neurologic outcome and inhospital mortality were comparable between high-volume (27.7% and 44.6%, respectively) and low-volume hospitals (21.1% and 40.5%) in the propensity-matched cohorts. The adjusted odds ratio for the high-volume hospitals compared with low-volume hospitals was 1.506 (95% confidence interval, 0.875-2.592) for poor neurologic outcome. Conclusions: Higher TTM case volume was significantly associated with early initiation of TTM and lower incidence of adverse events. However, case volume had no association with neurologic outcome and inhospital mortality. © 2014 Elsevier Inc. All rights reserved.

1. Introduction Targeted temperature management (TTM) during the initial hospital period after resuscitation for an out-of-hospital cardiac arrest (OHCA) is thought to be a critical component of integrated postcardiac arrest care. Two landmark clinical trials have demonstrated that induction of mild hypothermia (32°C to 34°C) is associated with improved survival and neurologic outcomes in comatose survivors resuscitated from ventricular fibrillation [1,2]. Case volume is important for the optimization of hospital performance at certain procedures. Higher case volume is associated with maintenance of individual knowledge and skills and enables a more

☆ Funding sources/disclosures: The authors have no relevant financial information or potential conflicts of interest to disclose. ⁎ Corresponding author. Tel.: +82 62 220 6809; fax: +82 62 228 7417. E-mail address: [email protected] (B.K. Lee).

efficient delivery of multidisciplinary team-based treatment. Volumeoutcome relationships have been studied in a number of settings. Patients treated at high-volume hospitals have better outcomes than those treated at low-volume hospitals [3,4]. In addition, volumeoutcome relationships are evident for medical conditions and procedures that require a highly complex, multidisciplinary approach [5-9]. Targeted temperature management after resuscitation from cardiac arrest is a resource- and labor-intensive treatment that requires a team approach and adherence to protocol. Therefore, case volume, which can be used as a surrogate marker of quality of care, may play an important role in the outcome of TTM. The relationship between hospital factors and outcomes of postcardiac arrest patients has been evaluated previously [6,7]. However, no study has evaluated directly the effect of case volume on TTM performance and outcome, despite the importance of TTM on the recovery of comatose cardiac arrest victims. Thus, in the present study, we sought to determine the effect of case volume on TTM performance, incidence of adverse events, and outcomes in comatose cardiac arrest survivors treated with TTM. We hypothesized that hospitals with higher

http://dx.doi.org/10.1016/j.ajem.2014.10.003 0735-6757/© 2014 Elsevier Inc. All rights reserved.

Please cite this article as: Lee SJ, et al, Impact of case volume on outcome and performance of targeted temperature management in out-of-hospital cardiac arrest survivors, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.10.003

2

S.J. Lee et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx

case volume have better TTM performance, lower incidence of adverse events, lower inhospital mortality, and better neurologic outcomes than hospitals with low case volumes.

The study was approved by the institutional review board of each of the 24 hospitals contributing to the KORHN study registry. 2.1. Data source

2. Methods A retrospective observational cohort study was performed using data from the Korean Hypothermia Network (KORHN) study registry.

A total of 24 hospitals throughout the Republic of Korea participated in the KORHN study registry, a Web-based, multicenter registry of adult (≥ 18 years) comatose OHCA patients treated with TTM (Figure and

Figure. Geographic distribution of the 24 participating hospitals across the Republic of Korea.

Please cite this article as: Lee SJ, et al, Impact of case volume on outcome and performance of targeted temperature management in out-of-hospital cardiac arrest survivors, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.10.003

S.J. Lee et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx Table 1 Characteristics of the 24 participating hospitals Geography Metropolitan Nonmetropolitan Affiliation with medical school Hospital level Tertiary referral hospital Secondary hospital Beds N1500 beds 1001-1500 beds 501-1000 beds b500 beds ED levela Level 1 Level 2 Level 3 Emergency physician, median (IQR) Coronary interventionist, median (IQR) ICU beds, median (IQR) Nurse-to-patient ratio 1:2 1:3 1:4

17 (70.8) 7 (29.2) 23 (95.8) 18 (75.0) 6 (25.0) 3 (12.5) 4 (16.7) 15 (62.5) 2 (8.3) 9 (37.5) 15 (62.5) 0 (0.0) 6 (5-7) 5 (4-6) 19 (9-25) 5 (20.8) 16 (66.7) 3 (12.5)

Abbreviation: ICU, intensive care unit. a Emergency department level, all EDs in Korea are designated as levels 1 to 3 based on the human resource, equipments, and instruments by law.

Table 1). Briefly, an investigator at each hospital was trained to correctly extract data from hospital medical records and transcribe the data into the standardized report form used to enter information about each case treated with TTM. The investigators reviewed hospital medical records of all comatose cardiac arrest patients treated with TTM and entered the data into the registry. Three clinical research associates checked the data and confirmed the data by sending queries to investigators. Finally, a data manager rechecked the data and decided whether the data would be accepted or required revision. Between January 2007 and December 2012, 930 adult comatose OHCA survivors treated with TTM were included in the registry. For all hospitals, the TTM regimen was cooling to 32°C to 34°C for 12 to 24 hours. The participating hospitals treated the patients according to their own treatment protocol, and TTM was performed with any kind of temperature management equipment. 2.2. Study patients and variables Adult OHCA survivors (≥18 years) who were treated with TTM at the participating hospitals between January 2007 and December 2012 were included in the study. Patients were excluded if the cardiac arrest resulted from trauma or if the prearrest cerebral performance category (CPC) score was greater than or equal to 3. The following variables were extracted from the KORHN study registry database: age, sex, preexisting diseases, CPC score before arrest, location before emergency department (ED) admission (the patient was transferred from another institution or was transported from the scene of collapse), presence of a witness on collapse, bystander cardiopulmonary resuscitation (CPR), etiology of cardiac arrest (cardiac or extracardiac), coronary angiography, first monitored rhythm (shockable, nonshockable, or unknown), time from collapse to restoration of spontaneous circulation (ROSC), initial glucose after ROSC, Glasgow Coma Scale (GCS) after ROSC, time from ROSC to initiation of cooling, time from initiation of cooling to achieve target temperature, duration of cooling maintenance, rewarming duration, cooling method (automated device with temperature feedback control or conventional method), vital status at hospital discharge (alive or dead), and CPC at hospital discharge. Data on the following adverse events during TTM were extracted: unintentional overcooling (b 32°C), bradycardia (b40 beats/min), hypokalemia (b 3.0 mEq/L), hyperkalemia (N5.0 mEq/L), hypoglycemia (b 80 mg/dL), hyperglycemia (N180 mg/dL), bleeding, hypotension (systolic blood

3

pressure b90 mm Hg or mean arterial pressure b60 mm Hg for at least 30 minutes; or need for supportive measures to maintain systolic blood pressure N90 mm Hg or mean arterial pressure N 60 mm Hg), seizure (clinically involuntary movement or epileptiform discharge on electroencephalogram), and rebound hyperthermia (N 38°C after rewarming). 2.3. Volume thresholds Sensitivity analysis was performed to determine the cutoff between low volume and high volume [10]. Sensitivity analysis used the area under receiver operating characteristic curve for poor neurologic outcome at hospital discharge (CPC 3-5), according to the annual number of cases for each hospital. The case volume of each hospital was calculated as the mean annual number of TTM episodes performed at each hospital during the study period. The cutoff number was defined as the trade-off between sensitivity and specificity. 2.4. Propensity score matching To determine the effects of volume on outcomes, propensity scoring was used to match patients at high-volume hospitals to patients at lowvolume hospitals, to determine the effects of volume on the outcomes. Propensity scores were calculated using a logistic regression model. The dependent variable was treatment at a high-volume hospital or low-volume hospital, and the independent variables (covariates) were the potential risk factors of age, sex, preexisting diseases, presence of a witness on collapse, bystander CPR, etiology of cardiac arrest, first monitored rhythm, time from collapse to ROSC, initial glucose after ROSC, GCS after ROSC, and location before ED admission. Propensity scores were categorized into 10 decimals. Using the estimated propensity score, patients treated in low-volume hospitals were matched to patients treated in high-volume hospitals whose estimated propensity score was within 0.6 SDs, which has been shown to eliminate approximately 90% of the bias [11]. 2.5. Outcome The primary outcome was neurologic outcome, which was assessed using CPC at hospital discharge according to the recommendations for outcome assessment in comatose cardiac arrest patients and was recorded as CPC 1 (good performance), CPC 2 (moderate disability), CPC 3 (severe disability), CPC 4 (vegetative state), or CPC 5 (brain death or death) [12]. Neurologic outcome was dichotomized as either good (CPC 1 and CPC 2) or poor (CPC 3-5). The secondary outcomes were inhospital mortality, TTM performance, and the incidence of adverse events during TTM. Targeted temperature management performance was assessed using time from ROSC to initiation of cooling, time from initiation of cooling to achieve target temperature, duration of cooling maintenance, and rewarming duration. 2.6. Statistical analyses Continuous variables are presented as median values and interquartile ranges (IQRs). The continuous variables showed nonnormal distribution; thus, the Mann-Whitney U test was performed to compare continuous variables. Categorical variables are presented as frequencies and percentages. Categorical variables were compared using the χ 2 test or Fisher exact test. Multivariate binary logistic regression analysis was used to assess independent risk factors on poor neurologic outcome. All variables with a significance of P b .1 in the univariate analyses were included in the multivariate logistic regression model. Collinearity between variables was excluded before modeling. Backward selection was used to reach the final model. Goodness of fit of the final model was evaluated using the Hosmer-Lemeshow test. Data were analyzed using PASW/SPSS software, version 18 (IBM, Chicago, IL). Significance was set at P b .05.

Please cite this article as: Lee SJ, et al, Impact of case volume on outcome and performance of targeted temperature management in out-of-hospital cardiac arrest survivors, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.10.003

4

S.J. Lee et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx

Table 2 Clinical characteristics according to neurologic outcome at hospital discharge Total (n = 901) Age, y Male sex Preexisting diseases Coronary artery disease Congestive heart failure Hypertension Diabetes Pulmonary disease Renal impairment Stroke Liver cirrhosis Malignancy Witness of collapse Bystander CPR Etiology of cardiac arrest Cardiac First monitored rhythm Shockable Nonshockable Unknown Time from collapse to ROSC, min Initial glucose after ROSC, mg/dL GCS after ROSC Location before ED admission From the scene of collapse From another institution Time from ROSC to initiation of cooling, min Time from initiation of cooling to achieving target temperature, min Duration of cooling maintenance, min Rewarming duration, min Cooling method Conventional method Automated device

Good (n = 248)

Poor (n = 653)

P

57 (46-70) 628 (69.7)

50.0 (40.0-60.0) 193 (77.8)

62.0 (49.0-72.0) 435 (66.6)

b.001 .001

791 (87.8) 872 (96.8) 593 (65.8) 701 (77.8) 848 (94.1) 848 (94.1) 860 (95.4) 891 (98.9) 877 (97.3) 598 (66.4) 274 (30.4)

34 (13.7) 5 (2.0) 65 (26.2) 31 (12.5) 4 (1.6) 4 (1.6) 5 (2.0) 1 (0.4) 5 (2.0) 202 (81.5) 100 (40.3)

76 (11.6) 24 (3.7) 243 (37.2) 169 (25.9) 49 (7.5) 49 (7.5) 36 (5.5) 9 (1.4) 19 (2.9) 396 (60.7) 174 (26.6)

.396 .208 .002 b.001 .001 .001 .024 .212 .457 b.001 b.001

554 (61.5)

220 (88.7)

334 (51.1)

b.001 b.001

229 (25.4) 638 (70.8) 34 (3.8) 31.0 (22.0-42.0) 234.0 (172.5-313.0) 3 (3-3)

142 (57.7) 91 (37.0) 13 (5.3) 25.0 (17.0-35.0) 226.5 (166.0-279.8) 3 (3-5)

87 (13.4) 546 (84.4) 14 (2.2) 33.0 (25.0-45.0) 250.0 (176.0-329.0) 3 (3-5)

623 (69.1) 278 (30.9) 101.0 (46.0-203.0) 156.0 (80.0-270.0) 1440 (1290-1440) 690 (420-900)

155 (62.5) 93 (37.5) 118.0 (56.3-222.3) 195.0 (120.0-330.0) 1440 (1320-1440) 660 (420-780)

468 (71.7) 185 (28.3) 96.5 (43.0-192.8) 135.0 (65.5-240.0) 1440 (1288-1440) 720 (444-960)

163 (18.1) 738 (81.9)

24 (9.7) 224 (90.3)

b.001 .001 b.001 .008

.040 b.001 .280 .009 b.001

139 (21.3) 514 (78.7)

Data are presented as n (%) or medians with IQRs.

3. Results 3.1. Demographic findings During the study period, 930 adult OHCA survivors underwent TTM at the participating hospitals. Of these patients, 29 were excluded for prearrest CPC greater than or equal to 3. A total of 901 patients were included in this study. Of the patients included in the analysis, 544 patients (60.4%) survived to hospital discharge, and 247 patients (27.5%) showed good neurologic outcome at hospital discharge. Clinical characteristics of the included patients are shown in Table 2. 3.2. Sensitivity analysis and propensity score matching The institutional TTM case volume ranged from 2 to 29.5 per year (median, 7.2; IQR, 5.0-14.6) for the participating hospitals. The cutoff value was calculated to be 15.5 per year using sensitivity analysis for neurologic outcome, and sensitivity and specificity were 52.1% and 61.4%, respectively. A total of 437 patients (48.5%) were treated at low-volume hospitals, and 464 patients (51.5%) were treated at highvolume hospitals. The clinical characteristics of patients treated at high- and low-volume hospitals are shown in Table 3. Patients at high-volume hospitals were significantly younger and had a lower rate of previous stroke, a higher rate of cardiac etiology, a higher rate of shockable rhythm, a higher GCS, a higher rate of transfer from another hospital, longer time from ROSC to initiation of cooling, longer cooling maintenance, shorter rewarming duration, and a higher rate of good neurologic outcome. Coronary angiography was performed in 86 (34.3% of 251 patients with cardiac etiology) patients at lowvolume hospitals and in 140 (46.2% of 303 patients with cardiac etiology) patients at high-volume hospitals (P = .004). After propensity score matching, 289 patients were assigned to each group. There were no significant differences in the distribution of

potential risk factors between the 2 groups after propensity score matching (Table 3). High-volume hospitals had shorter time from ROSC to initiation of cooling; however, no significant differences were observed between the 2 groups for the time from initiation of cooling to achieving target temperature, the duration of cooling maintenance, and the rewarming duration. Coronary angiography was not associated with case volume (57 of 165 patients with cardiac etiology at lowvolume hospitals; 69 of 179 patients with cardiac etiology at highvolume hospitals; P = .441). Neurologic outcome at hospital discharge and inhospital mortality did not differ between the 2 groups. 3.3. Association of case volume with neurologic outcome at hospital discharge After propensity score matching, multivariate logistic regression analysis showed that age, etiology of cardiac arrest, time from collapse to ROSC, GCS after ROSC, and first monitored rhythm were independent risk factors for poor neurologic outcome (Table 4). Case volume had no effect on the neurologic outcome. 3.4. Adverse events during TTM The adverse events during TTM in the propensity score–matched cohort are shown in Table 5. The low-volume hospitals had higher rates of hyperglycemia, bleeding, and hypotension. In addition, patients treated at the low-volume hospitals were more likely to have rebound hyperthermia during rewarming. 4. Discussion To our knowledge, this is the first study to examine the relationship between case volume and TTM performance, adverse event incidence, and outcome. In this propensity score–matched analysis of retrospective

Please cite this article as: Lee SJ, et al, Impact of case volume on outcome and performance of targeted temperature management in out-of-hospital cardiac arrest survivors, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.10.003

S.J. Lee et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx

5

Table 3 Clinical characteristics of unmatched and propensity score–matched cohorts according to case volume Unmatched cohort Low volume (n=437) Age, y Male sex Preexisting diseases Coronary artery disease Congestive heart failure Hypertension Diabetes Pulmonary disease Renal impairment Stroke Liver cirrhosis Malignancy Witness of collapse Bystander CPR Etiology of cardiac arrest Cardiac First monitored rhythm Shockable Nonshockable Unknown Time from collapse to ROSC, min Initial glucose after ROSC, mg/dL GCS after ROSC Location before ED admission From the scene of collapse From another institution Time from ROSC to initiation of cooling, min Time from initiation of cooling to achieving target temperature, min Duration of cooling maintenance, min Rewarming duration, min Cooling method Conventional method Automated device Good neurologic outcome at hospital discharge Inhospital mortality

Propensity score–matched cohort High volume (n=464)

P

Low volume (n = 289)

High volume (n = 289)

P

59.0 (48.0-71.0) 295 (67.5)

56.0 (46.0-69.0) 333 (71.8)

.030 .164

58.0 (48.0-69.5) 200 (69.2)

58.0 (47.0-70.0) 201 (69.6)

47 (10.8) 14 (3.2) 154 (35.2) 88 (20.1) 24 (5.5) 26 (5.9) 27 (6.2) 4 (0.9) 12 (2.7) 278 (63.6) 135 (30.9)

63 (13.6) 15 (3.2) 154 (33.2) 112 (24.1) 29 (6.3) 27 (5.8) 14 (3.0) 6 (1.3) 12 (2.6) 320 (69.1) 139 (30.0)

.196 .980 .517 .149 .629 .934 .023 .589 .882 .081 .760

26 (9.0) 7 (2.4) 107 (37.0) 51 (17.6) 18 (6.2) 13 (4.5) 19 (6.6) 3 (1.0) 5 (1.7) 188 (65.1) 81 (28.0)

34 (11.8) 10 (3.5) 99 (34.3) 68 (23.5) 17 (5.9) 19 (6.6) 13 (4.5) 4 (1.4) 8 (2.8) 195 (67.7) 94 (32.5)

.276 .460 .487 .080 .862 .275 .275 1.000 .400 .499 .239

251 (57.4)

303 (65.3)

.015 .001

165 (57.1)

179 (61.9)

.235 .334

64 (22.1) 216 (74.7) 9 (3.1) 31.5 (23.8-44.3) 243 (161-311) 3 (3-3)

78 (27.0) 200 (69.2) 11 (3.8) 31.0 (23.0-42.0) 251 (181-315) 3 (3-3)

231 (79.9) 58 (20.1) 90.0 (45.0-174.0) 150.0 (70.0-250.0) 1380 (1200-1440) 709 (413-1020)

247 (85.5) 42 (14.5) 63.5 (31.0-124.5) 170.0 (80.0-300.0) 1440 (1290-1440) 660 (480-780)

93 (21.3) 334 (77.0) 10 (2.3) 32.0 (23.0-44.0) 239.5 (167.0-312.0) 3 (3-3)

136 (29.3) 304 (65.6) 24 (5.2) 30.0 (22.0-41.0) 245.0 (177.3-314.8) 3 (3-3)

375 (85.5) 62 (14.2) 91 (45.0-172.0) 150 (70-250) 1380 (1214-1442) 720 (420-1058)

248 (53.4) 216 (46.6) 113.5 (49.0-235.0) 170 (90-280) 1440 (1380-1440) 660 (450-780)

87 (19.9) 350 (80.1) 96 (22.0) 182 (41.6)

76 (16.4) 388 (83.6) 152 (32.8) 175 (37.7)

.316 .558 b.001 b.001

.009 .069 b.001 .018 .169

b.001 .228

55 (19.0) 234 (81.0) 61 (21.1) 117 (40.5)

.847 .928

.638 .493 .276 .079

.001 .304 .058 .267 .130

70 (24.2) 219 (75.8) 80 (27.7) 129 (44.6)

.066 .313

Data are presented as n (%) or medians with IQRs.

data from 24 hospitals in Korea, higher case volume was significantly associated with early initiation of cooling and lower incidence of adverse events. However, case volume had no effect on neurologic outcome, despite the association of case volume with TTM performance and adverse event incidence. Several studies have found an association between procedural volume for complex procedures and procedure performance and clinical outcome [3,4,13]. In a study evaluating primary percutaneous coronary intervention performance in England and Wales, high-volume centers had lower door-to-balloon times [13]. Consistent with these studies, in the present study, hospitals with higher TTM case volumes had shorter times from ROSC to initiation of cooling. The time to initiation of cooling in the present study reflects hospital-based performance of TTM because prehospital cooling is not available in Korea. Several studies have shown that higher case volume is also associated with fewer procedural complications [14,15]. In a study of the relationship between

Table 4 Independent predictors of unfavorable neurologic outcome in propensity score– matched cohorts

Age Cardiac etiology Time from collapse to ROSC GCS Shockable rhythm High volume

OR (95% CI)

P

1.053 (1.033-1.074) 0.099 (0.045-0.217) 1.075 (1.053-1.098) 0.668 (0.507-0.880) 0.234 (0.129-0.424) 1.506 (0.875-2.592)

b.001 b.001 b.001 .004 b.001 .139

Abbreviations: OR, odds ratio; CI, confidence interval.

procedural volume and complication rate after implantable cardioverterdefibrillator placement using the National Cardiovascular Data Registry's ICD Registry, the complication rate decreased with increasing procedural volume [15]. Similarly, in the present study, high-volume hospitals had lower incidences of hyperglycemia, bleeding, hypotension, and rebound hyperthermia than the low-volume hospitals, indicating that case volume is important in avoiding adverse events in patients undergoing TTM. Several studies in animals have indicated that the earlier the cooling is initiated after cardiac arrest, the better the outcome [16,17]. Although the association between rapid time to cooling and positive outcomes has not been confirmed in humans, several clinical studies suggest that this association exists in humans [18,19]. In a study that investigated the association of adverse events during TTM with mortality in 765 patients

Table 5 Adverse events during targeted temperature management in propensity score– matched cohort

Overcooling Bradycardia Hypokalemia Hyperglycemia Bleeding Hypotension Seizure Hyperthermia Hyperkalemia Hypoglycemia

Low volume (n = 289)

High volume (n = 289)

P

70 (24.2) 37 (12.8) 70 (24.2) 173 (59.9) 19 (6.6) 126 (43.6) 97 (33.6) 39 (13.5) 20 (6.9) 27 (9.3)

53 (18.3) 53 (18.3) 82 (28.4) 103 (35.6) 8 (2.8) 85 (29.4) 78 (27.0) 24 (8.3) 15 (5.2) 22 (7.6)

.084 .066 .257 b.001 .030 b.001 .085 .045 .383 .455

Data are presented as n (%).

Please cite this article as: Lee SJ, et al, Impact of case volume on outcome and performance of targeted temperature management in out-of-hospital cardiac arrest survivors, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.10.003

6

S.J. Lee et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx

from 22 centers, seizure and sustained hyperglycemia were independently associated with increased mortality [20]. In the present study, high-volume hospitals had significantly shorter times to initiation of cooling and significantly lower incidences of hyperglycemia in the propensity score–matched cohorts; however, case volume had no effect on the outcomes. There are several possible explanations for the lack of association between case volume and outcomes observed in the present study. First, the study may have had limited power to detect differences; the number of hospitals studied was relatively small (n = 24). Second, there was a low rate of bystander CPR and a long duration between collapse and ROSC in our patients. The time to the initiation of cooling or the incidence of hyperglycemia may be unrelated to outcomes in the presence of other more important factors, such as low rates of bystander CPR or long duration between collapses and ROSC [21]. Third, a relatively high proportion of patients who presented with nonshockable rhythm (N70%) were included in the study. In several studies on OHCA patients treated with TTM, cooling was not associated with good outcomes in patients presenting with nonshockable rhythm but was independently associated with improved outcomes in patients with shockable rhythm [22,23]. Fourth, the lack of association between case volume and outcomes despite earlier initiation of cooling in the high-volume hospitals may reflect the fact that it is not the cooling that matters in postcardiac arrest care. The TTM regimen in the present study was cooling at 32°C to 34°C for 12 to 24 hours. A recent randomized controlled trial by Nielsen et al [24] compared 2 target temperatures in comatose OHCA survivors and found that cooling at a targeted temperature of 33°C did not confer a benefit on mortality or neurologic outcome when compared with a targeted temperature of 36°C. There are potential limitations to this study. First, the study was a retrospective, observational study, although it used propensity score matching to compare between high- and low-volume hospitals. Despite adjusting for potential confounders, there may have been residual unmeasured confounders. Second, the participating hospitals were not randomly selected because participation in the KORHN registry is voluntary. In Korea, TTM in postcardiac arrest care is limited mostly to large, university-affiliated hospitals. Thus, this study was conducted in 24 hospitals with similar characteristics and included 1 large district general hospital and 23 university-affiliated teaching hospitals. The nonrandom selection of the participating hospitals could have influenced the findings. Third, this study involved a small number of hospitals in a single country. Therefore, generalizability to other country settings is uncertain. Fourth, the calculated power of association between primary outcome and case volume was 0.455 despite the use of data obtained from a multicenter registry. Thus, this study did not have adequate power to precisely assess the effects of case volume. Finally, we considered only TTM case volumes in our study. Several hospital factors other than case volume, including critical care physician volume, nurse-to-patient ratio, and intensive care protocol, may have significant impact on outcomes. Patients resuscitated from OHCA require multiple, time-sensitive interventions that are continuously available. However, TTM after cardiac arrest is currently available in only relatively few hospitals in many countries, including Korea [25,26]. Considering the described limitations, the present study cannot be considered to be confirmatory evidence for the absence of an association between TTM case volume and outcome. A prospective study that includes a larger number of hospitals with information on various process measures is needed to provide further insight into volume-outcome relationships in TTM after cardiac arrest. Nevertheless, our finding that case volume had no effect on the outcomes, together with the time-sensitive nature of postcardiac arrest care and the underutilization of TTM, indicates that health policy that focuses on implementing TTM into more hospitals, rather than limiting TTM to select large referral hospitals, may have benefits in improving outcomes after OHCA.

5. Conclusions In this propensity score matching analysis of retrospective data from 24 hospitals in Korea, higher case volume was significantly associated with early initiation of TTM and lower incidence of adverse events. However, case volume had no effect on the neurologic outcome. References [1] Bernard SA, Gray TW, Buist MD, Jones BM, Silvester W, Gutteridge G, et al. Treatment of comatose survivors of out-of-hospital cardiac arrest with induced hypothermia. N Engl J Med 2002;346:557–63. [2] Hypothermia after Cardiac Arrest Study G. Mild therapeutic hypothermia to improve the neurologic outcome after cardiac arrest. N Engl J Med 2002;346:549–56. [3] Finks JF, Osborne NH, Birkmeyer JD. Trends in hospital volume and operative mortality for high-risk surgery. N Engl J Med 2011;364:2128–37. [4] Kuwabara H, Fushimi K, Matsuda S. Relationship between hospital volume and outcomes following primary percutaneous coronary intervention in patients with acute myocardial infarction. Circ J 2011;75:1107–12. [5] Canto JG, Every NR, Magid DJ, Rogers WJ, Malmgren JA, Frederick PD, et al. The volume of primary angioplasty procedures and survival after acute myocardial infarction. National Registry of Myocardial Infarction 2 Investigators. N Engl J Med 2000;342:1573–80. [6] Carr BG, Goyal M, Band RA, Gaieski DF, Abella BS, Merchant RM, et al. A national analysis of the relationship between hospital factors and post-cardiac arrest mortality. Intensive Care Med 2009;35:505–11. [7] Carr BG, Kahn JM, Merchant RM, Kramer AA, Neumar RW. Inter-hospital variability in post-cardiac arrest mortality. Resuscitation 2009;80:30–4. [8] Dudley RA, Johansen KL, Brand R, Rennie DJ, Milstein A. Selective referral to highvolume hospitals: estimating potentially avoidable deaths. JAMA 2000;283:1159–66. [9] Nathens AB, Jurkovich GJ, Maier RV, Grossman DC, MacKenzie EJ, Moore M, et al. Relationship between trauma center volume and outcomes. JAMA 2001;285:1164–71. [10] Shin SD, Suh GJ, Ahn KO, Song KJ. Cardiopulmonary resuscitation outcome of out-ofhospital cardiac arrest in low-volume versus high-volume emergency departments: an observational study and propensity score matching analysis. Resuscitation 2011;82:32–9. [11] Gu XS, Rosenbaum PR. Comparison of multivariate matching methods: structures, distances, and algorithms. J Comput Graph Stat 1993;2:405–20. [12] Booth CM, Boone RH, Tomlinson G, Detsky AS. Is this patient dead, vegetative, or severely neurologically impaired? Assessing outcome for comatose survivors of cardiac arrest. JAMA 2004;291:870–9. [13] West RM, Cattle BA, Bouyssie M, Squire I, de Belder M, Fox KA, et al. Impact of hospital proportion and volume on primary percutaneous coronary intervention performance in England and Wales. Eur Heart J 2011;32:706–11. [14] Al-Khatib SM, Lucas FL, Jollis JG, Malenka DJ, Wennberg DE. The relation between patients' outcomes and the volume of cardioverter-defibrillator implantation procedures performed by physicians treating Medicare beneficiaries. J Am Coll Cardiol 2005;46:1536–40. [15] Freeman JV, Wang Y, Curtis JP, Heidenreich PA, Hlatky MA. Physician procedure volume and complications of cardioverter-defibrillator implantation. Circulation 2012;125:57–64. [16] Abella BS, Zhao D, Alvarado J, Hamann K, Vanden Hoek TL, Becker LB. Intra-arrest cooling improves outcomes in a murine cardiac arrest model. Circulation 2004; 109:2786–91. [17] Nozari A, Safar P, Stezoski SW, Wu X, Kostelnik S, Radovsky A, et al. Critical time window for intra-arrest cooling with cold saline flush in a dog model of cardiopulmonary resuscitation. Circulation 2006;113:2690–6. [18] Mooney MR, Unger BT, Boland LL, Burke MN, Kebed KY, Graham KJ, et al. Therapeutic hypothermia after out-of-hospital cardiac arrest: evaluation of a regional system to increase access to cooling. Circulation 2011;124:206–14. [19] Wolff B, Machill K, Schumacher D, Schulzki I, Werner D. Early achievement of mild therapeutic hypothermia and the neurologic outcome after cardiac arrest. Int J Cardiol 2009;133:223–8. [20] Nielsen N, Sunde K, Hovdenes J, Riker RR, Rubertsson S, Stammet P, et al. Adverse events and their relation to mortality in out-of-hospital cardiac arrest patients treated with therapeutic hypothermia. Crit Care Med 2011;39:57–64. [21] Nielsen N, Hovdenes J, Nilsson F, Rubertsson S, Stammet P, Sunde K, et al. Outcome, timing and adverse events in therapeutic hypothermia after out-of-hospital cardiac arrest. Acta Anaesthesiol Scand 2009;53:926–34. [22] Dumas F, Grimaldi D, Zuber B, Fichet J, Charpentier J, Pene F, et al. Is hypothermia after cardiac arrest effective in both shockable and nonshockable patients?: insights from a large registry. Circulation 2011;123:877–86. [23] Oddo M, Schaller MD, Feihl F, Ribordy V, Liaudet L. From evidence to clinical practice: effective implementation of therapeutic hypothermia to improve patient outcome after cardiac arrest. Crit Care Med 2006;34:1865–73. [24] Nielsen N, Wetterslev J, Cronberg T, Erlinge D, Gasche Y, Hassager C, et al. Targeted temperature management at 33°C versus 36°C after cardiac arrest. N Engl J Med 2013;369:2197–206. [25] Abella BS, Rhee JW, Huang KN, Vanden Hoek TL, Becker LB. Induced hypothermia is underused after resuscitation from cardiac arrest: a current practice survey. Resuscitation 2005;64:181–6. [26] Patel PV, John S, Garg RK, Temes RE, Bleck TP, Prabhakaran S. Therapeutic hypothermia after cardiac arrest is underutilized in the United States. Ther Hypothermia Temp Manag 2011;1:199–203.

Please cite this article as: Lee SJ, et al, Impact of case volume on outcome and performance of targeted temperature management in out-of-hospital cardiac arrest survivors, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.10.003