Article, Cardiology

The effect of interhospital transfers, emergency medical services, and distance on ischemic time in a rural ST-elevation myocardial infarction system of care

a b s t r a c t

Background: Regional myocardial infarction systems of care have been shown to improve timely access to Primary percutaneous coronary intervention . However, there is a relatively sparse research on rural “frontier” regions. Arrival mode, high rates of Interhospital transfers, long transport times, low population densi- ty, and mostly volunteer emergency medical services (EMS) distinguish this region from metropolitan systems of care. We sought to assess the effect of interhospital transfers, distance, and arrival mode on total Ischemic times for patients with ST-elevation myocardial infarctions undergoing Primary PCI.

Methods: We assessed patient data from our observational cohort of 395 patients with ST-elevation myocardial infarction with PCI as their primary treatment strategy. Data came from the 10 PCI hospitals participating in the Wyoming Mission: Lifeline program from January 2013 to September 2014. We performed both regression and tests of differences.

Results: Median total ischemic time was nearly 2.7 times greater in Transferred patients than those presenting di- rectly (379 vs 140 minutes). Distance in miles traveled between patient’s home and PCI facility was 2.5 times larger in transfer patients (51 vs 20 miles). Emergency medical services arrival was associated with 23% shorter total ischemic times than self-arrival.

Conclusions: Transfer patients from referral hospitals had significantly greater total ischemic time, and use of EMS was associated with significantly lower times. Transport distance was mixed in its effect. These findings suggest a continued focus on improving transitions between referral and receiving centers and enhancing coordination in rural systems of care to reduce the multiplier effect of transfers on total ischemic time.

(C) 2015

Introduction

Percutaneous coronary intervention (PCI) is the preferred treatment strategy for ST-elevation myocardial infarction , to increase myocardial salvage and decrease mortality rates [1]. Door to balloon (D2B) and total ischemic time (measured as symptom onset to arterial reperfusion, or SOAR) are important quality outcome measures, although in practice, less than one half of patients with STEMI receive reperfusion within the guideline-recommended 90-minute D2B time [2]. Among transfer patients at PCI facilities, this percentage drops to only 4.2% of patients receiving reperfusion within the recommended D2B time of 90 minutes or less [3].

* Corresponding author at: University of Texas Health Sciences Center, 6431 Fannin Street, JJL-450 E, Houston, TX 77030. Tel.: +1 713 500 7836; fax: +1 713 500 7884.

E-mail address: [email protected] (J.R. Langabeer).

In Rural settings, the recommended time goals are more difficult to achieve, yet very little has been researched on rural systems [4]. Among frontier states like Wyoming, treatment times are difficult given the unique challenges resulting from low population density, large geographic distances, long transport times, and mostly volun- teer emergency medical services (EMS). A recent study by Langabeer et al [5] examined state by state growth of PCI capacity and reported that Wyoming had the third fewest PCI facilities per capita, nearly half that of most other states (approximately 3.7 facilities for every million in population vs national average of 6.98). Wyoming has the second lowest PCI density in the nation on a per square mile basis, with 2.04 PCI hospitals per 100 000 square miles. Although Wyoming has a lower prevalence of myocardial infarction (32 per 1000 persons), the large geographic distances make it less likely for patients to receive appropriate care within the recommended time to treatment window.

http://dx.doi.org/10.1016/j.ajem.2015.04.009

0735-6757/(C) 2015

Treatment protocols“>914 J.R. Langabeer II et al. / American Journal of Emergency Medicine 33 (2015) 913916

The aim of this study was to evaluate the temporal impact of inter- hospital transfers on quality outcomes for patients with STEMI undergo- ing primary PCI in rural Wyoming.

Methods

Program design/treatment protocols

The state of Wyoming has an estimated population of 582658. Al- though the 3 PCI-capable hospitals are distributed throughout all parts of the state (and 7 in bordering states), the rural terrain presents unique challenges for patients with STEMI requiring transfer to a STEMI- accepting facility. In 2013, the American Heart Association (AHA) established Mission: Lifeline Wyoming, a 3-year, $5.9 million initiative funded by The Leona M. and Harry B. Helmsley Charitable Trust to the Southwest Affiliate of AHA. In an effort to improve systems of care and reduce treatment time for patients with STEMI, the initiative involves collaboration among EMS providers, First responders, and hospitals.

These 10 STEMI-receiving facilities accept transfer patients from ap- proximately 25 STEMI-referring acute and critical access hospitals. A guideline-based protocol was developed and implemented within the participating STEMI referral emergency departments. The protocol accounts for variation in arrival mode, specifically arrival by EMS, arrival to a STEMI-receiving center, or arrival to a STEMI-referring center. Once a patient with suspected STEMI was confirmed through 12-lead electro- cardiogram, protocols are in place for the emergency department and other necessary personnel. Multidisciplinary team meetings (consisting of participating EMS, state health department officials, STEMI-referring centers, and STEMI-receiving centers) were held to ensure ongoing evaluation of system outcomes. Protocols were developed, for Prehospital identification and activation and for receiving and transfer- ring PCI candidates.

Data and statistical analysis

Observational data from a total of 395 deidentified patients with STEMI who presented to the 10 Wyoming area for primary PCI were analyzed. Data were for all patients with STEMI who went through primary PCI from January 2013 to September 2014. Data were obtained from the hospitals directly on a quarterly basis, based on the standardized criteria for the National Cardiovascular Data Registry (NCDR) Action Registry. Nurse abstractors from the cardiac catheteriza- tion laboratory from each of the hospitals extracted their NCDR data and exported it in secure, encrypted files. These exports were identical to their registry data entered into NCDR. The file contained all patients with STEMI initially, although for this research, we focused only on STEMI primary PCI data. All direct identifiers were stripped from the data by the facilities before submission to AHA. A database was compiled in Microsoft Access (Microsoft; Redmond, WA) to store and manage all deidentified patient data. The study was approved by the institutional review board of the University of Texas Health Science Center.

The primary outcome assessed in this study was total ischemic time, defined as the time between symptom onset and arterial reperfusion (SOAR). We measured symptom onset as difference in time (in minutes) recorded between when the first device was activated in the catheteri- zation laboratory minus the time reported to the PCI facility, where the patient first noted Ischemic symptoms. We had missing data in some of these cases through listwise deletion. We used a logarithmic transformation of the dependent variable, SOAR, to reduce data skew- ness. We used the Breusch-Pagan/Cook-Weisberg test of variance to confirm that the heteroskedasticity assumptions were satisfied with the transformed variable.

We relied on analyses of variances to examine statistical differences in ischemic time, distances, and arrival patterns by groups. An Ordinary least squares regression model was developed to predict SOAR time after controlling for potential confounders. The primary independent

variables in the model tested the effects of whether the patient was transferred from an outside facility, distance from patient’s home zip code to a PCI hospital, and arrival mode. Distance was computed using geocoding of the centrode between the primary hospital latitude and longitude and the reported patient home zip code. The variables shock and cardiac arrest at first medical contact were included in the model to account for differences in patient presentation and case complexity. The model also controlled for patient-related differences for age and sex. Age and distance were included as continuous variables. The other factors in the model were dichotomous (yes/no). All probability values are 2-tailed and a value of P b .05 considered significant. Statistical analyses were performed using Stata statistics software (version 11.2,

StataCorp LP; College Station, TX).

Results

Data were merged and analyzed for the 21-month period, which in- cluded 711 total patients with STEMI presenting between January 1, 2013, and September 30, 2014. To ensure comparability of findings, we included in our statistical analyses only the subset of STEMI that underwent primary PCI (n = 395). We excluded 254 patients who had fibrinolysis therapy as their primary treatment strategy and 62 patients with contraindications to primary PCI. Therefore, this study focused ex- clusively on the effect of transfer, distance, and EMS on the patients with STEMI who underwent primary PCI (n = 395).

The patient characteristics and associated outcomes for the data in this study are shown in Table 1.

Significant variation in arrival patterns were noted in the Wyoming system during the study period. Wyoming is highly rural, and access to the nearest PCI hospital is an average distance of 58.2 miles for all pa- tients with STEMI in this study. Of patients arriving at PCI facilities, 42% were transfers from other facilities, either acute care or critical access, nearly double the national average. For transfer patients, the average dis- tance traveled from patient home to PCI facility was 72.6 vs 47.9 miles for nontransfers (median, 51.4 vs 20.3).

During this period, the mortality rate was 4.9% for patients with STEMI who underwent PCI. The median SOAR time was 206 minutes (interquartile range, 117-441 minutes), and the median D2B time was 56 minutes (interquartile range, 37-76 minutes). Total ischemic time was nearly 2.7 times greater for those patients who were transferred (379 vs 140 minutes). There were 438 (61.6%) patients who arrived to a PCI hospital by means of self-transport. Approximately 22% of patients presented in cardiac arrest at first medical contact and 7.9% in shock.

Using analysis of variance (ANOVA) testing, we found significance in

several areas. First, there was a significant difference in the total ischemic time differences for those who were transferred (379 minutes median) vs nontransferred (140 minutes) (P b .001). There was also a significant difference in the distance (mileage) for those patients transferred (20.3 vs 51.4 miles median) (P b .010). Those who arrived by EMS had SOAR

Table 1

Patient demographics and outcomes

Variable Patients with STEMI

primary PCI

Total patient volume, n 395

Age, y, mean, median, SD 64.1 +- 12.9

Male, n (%) 308 (78.0%)

Nonwhite, n (%) 15 (3.8%)

Self-transport, n (%) 245 (62.0%)

EMS transport, n (%) 150 (38.0%)

Transfer, n (%) 165 (41.8%)

Cardiac arrest on first medical contact, % of total 21.8%

Shock on first medical contact, % of total 7.9%

Mortality rate (%), unadjusted 4.9%

D2B, min, median (IQR) 54 (37-76)

Symptom onset to arterial reperfusion, min, median (IQR) 206 (117-441.5) Abbreviation: IQR, interquartile range.

J.R. Langabeer II et al. / American Journal of Emergency Medicine 33 (2015) 913916 915

times of 161 minutes median vs 256 minutes for those with patient- owned vehicle (P b .01). We had 23 cases of missing onset time, which explains the difference in the sample sizes below. Table 2 presents the differences in total ischemic time (SOAR) and distances by transfer vs nontransfer groups.

The ordinary least squares regression model confirmed the associa- tion of arrival mode, transfer, and sex with significantly lower SOAR times (R2 = .21, P b .001). When controlling for the other factors, dis- tance was not significant. In post hoc analyses, we computed interaction effect variables (Transfer x Distance, EMS x Transfer), and they were not significant; therefore, we show the final regression model in Table 3. To understand the temporal impact of the 2 significant variables on SOAR time, we exponentiated the coefficients of the log-transformed SOAR time. Accordingly, SOAR time is 138% higher for transfer than for nontransfer (median, 284 vs 206 minutes). Emergency medical services transport resulted in 23% reduced times than self-transport (47-minute

reduction) and 20% lower for male than for female (41 minutes).

Discussion

There is a somewhat disproportionate focus on urban STEMI systems of care in the United States and little understanding of the impact of transfers on rural time to treat outcomes. This study seeks to systematically assess the impact of transfers, distance, and arrival mode for patients presenting for primary PCI for STEMI in a rural frontier region of the United States with widely dispersed population. Our analysis included 395 patients with myocardial infarction in Wyoming over a 21-month period in 2013 through 2014. Several important observations can be made from our analyses.

Cardiovascular systems of care require collaboration among multiple groups to continuously improve quality of care and reduce time to treat. Multiple strategies have been documented to reduce delays, such as prehospital activation of the catheterization laboratory and electrocar- diogram transmission from the field. Community involvement and pub- lic awareness are integral parts of a successful STEMI system, even more so in rural areas characterized by long transportation times and patient transfer to PCI-capable facilities.

Transfers have a significant effect on total ischemic times in rural regions. In this study, we found that median total ischemic time was 138% greater in those patients who were transferred than those who presented directly to a PCI center (or 2.7 times greater in the ANOVA model). Although it is intuitive that transfers increase ischemic time, the precise estimate of the magnitude of this differential was not under- stood. After adjusting for pertinent variables, we found that high trans- fer rates were significantly associated with longer total ischemic times in both analyses of variances and regression. Consistent with other studies, we suggest that rural regions identify sources of delay and attempt to continuously remove barriers [6].

As we found here, EMS arrival to the hospital is significant in reduc- ing total ischemic time in rural environments. This appears similar in urban environments as well [7]. Emergency medical services activation of 38% for patients with STEMI is significantly lower in this rural region than more metropolitan areas, which might require public awareness campaigns or other patient outreach and education to improve SOAR times through greater EMS use. Mode of arrival remains important to reducing total ischemic time. Arrival by “self or patient own vehicle” re- sulted in significantly longer SOAR times (261 minutes for self-arrival vs

Table 2

Comparison of distance and SOAR in transfer vs nontransfer patients

Distance (miles) Total ischemic time (min)

Group

Mean

Median

n

IQR

Mean

Median

n

IQR

Nontransfer

47.89

20.30

230

2.9-92

281.52

140.0

219

98-226

Transfer

72.56

51.40

165

1.4-107

705.93

379.0

153

246-275

Total

58.22

23.26

395

1.8-101

456.08

206.0

372

117-442

Table 3

Regression results

Variable Coefficient SE P

Transfer

0.865

0.092

.000

Distance

0.000

0.000

.484

EMS arrival

-0.262

0.095

.006

Age

0.006

0.003

.091

Sex

-0.227

0.109

.038

Shock

-0.065

0.173

.707

Cardiac arrest

0.123

0.114

.279

161 minutes for EMS in the ANOVA model and 47-minute [23%] reduc- tion in the regression model), and therefore, arrival by EMS was associ- ated with significantly lower total times. Sex was also statistically significant in the final regression model, with males having 20% lower total ischemic times than females.

It is approximated that only 39% of hospitals in the United States are PCI-capable facilities [5]. However, these facilities are not geographically distributed equitably, so in certain areas such as Wyoming, the distance between patients and PCI-capable facilities is significant, averaging 58 miles between patient and PCI center in this study. This creates an even greater need for systematic collaboration and coordination of care between facilities because D2B and SOAR times on average will be higher than national estimates. We found that, although distance matters, when controlling for other confounding factors, it was not a significant influence on total ischemic time.

In this study, we focused on a more system-wide quality metric (SOAR) rather than D2B, which primarily measures only the PCI center treatment time. Total ischemic time has been referred to as the correct quality metric to focus on for STEMI care rather than the traditional D2B [8]. The value of using the SOAR variable as one of the outcomes in the study allows for analysis of systemic time from symptom onset to arterial reperfusion. We feel that our emphasis on total ischemic time and our focus on a statewide rural region (Wyoming) over such a long time horizon provide a significant contribution to the literature. Our analysis has limitations worth noting. First, this is an observa- tional study with inherent limitations and confounders that cannot be fully accounted by adjustments. Second, we are also unable to assess the impact of SOAR on clinical outcomes such as recurrent ischemia, re- peat revascularization, and survival. These we hope to further assess at a later date from the planned follow-up and subsequent analyses. How- ever, the relatively large-scale emphasis on rural regions and the state- wide collaborations in place between different health systems and the community provide for rich environment to assess STEMI systems of

care.

Conclusions

In conclusion, we demonstrate the significant time delay from inter- hospital transfers and the positive impact EMS can make on total ische- mic times in the rural state of Wyoming. The lessons learned from this rural, regional system of care suggest a continued focus on improving coordination in regional multistate systems. Similar programs can be implemented in other regions with low population density and broad geographic distances. These findings offer insight for other rural and frontier systems of care to develop effectively.

Acknowledgments

The authors thank the Wyoming Mission: Lifeline project, including the physicians, nurses, and EMS staff who have volunteered their time to the project. The authors also thank the funding from The Leona M. and Harry B. Helmsley Charitable Trust and AHA for their funding of the Wyoming Mission: Lifeline cardiovascular system of care.

916 J.R. Langabeer II et al. / American Journal of Emergency Medicine 33 (2015) 913916

References

  1. O’Gara PT, Kushner FG, Ascheim DD, Casey Jr DE, Chung MK, de Lemos JA, et al, DXCF/ AHA Task Force. 2013 ACCF/AHA guideline for the management of ST-elevation myo- cardial infarction: executive summary: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 2013;127:529-55.
  2. McNamara R, Herrin J, Bradley E, Portnay E, Curtis J, Wang Y, et al. Hospital improve- ment in time to reperfusion in patients with acute myocardial infarction, 1999 to 2002. J Am Coll Cardiol 2006;47(1):45-51.
  3. Nallamothu B, Bates E, Herrin J, Wang Y, Bradley E, Krumholz H. Times to treatment in transfer patients undergoing primary percutaneous coronary intervention in the United States: National Registry of Myocardial Infarction. Circulation 2005;111(6): 761-7.
  4. Aguirre F, Varghese J, Kelley M, Lam W, Lucore C, Gill J, et al. Rural interhospital trans- fer of ST-elevation myocardial infarction patients for percutaneous coronary revascu- larization: The Stat Heart Program. Circulation 2008;117:1145-52.
  5. Langabeer JR, Henry TD, Kereiakes DJ, DelliFraine J, Emert J, Wang Z, et al. Growth in per- cutaneous coronary intervention capacity relative to population and disease prevalence. J Am Heart Assoc 2013;2(6):1-7. http://dx.doi.org/10.1161/JAHA.113.000370.
  6. Miedema M, Newell M, Duval S, Garberich R, Handran C, Larson D, et al. Causes of delay and associated mortality in patients transferred with ST-segment-elevation myocardial infarc- tion. Circulation 2011(124):1-9. http://dx.doi.org/10.1161/CIRCULATIONAHA.111.033118.
  7. Bradley E, Herrin J, Wang Y, Barton B, Webster T, Mattera J, et al. Strategies for reducing the Door-to-balloon time in acute myocardial infarction. N Engl J Med 2006;355:2308-20.
  8. Denktas AE, Anderson HV, McCarthy J, Smalling RW. Total ischemic time: the correct focus of attention for optimal ST-segment elevation myocardial infarction care. J Am Coll Cardiol 2011;4(6).

Leave a Reply

Your email address will not be published. Required fields are marked *