Factors Influencing the Time to Administration of Thrombolytic Therapy With Recombinant Tissue Plasminogen

Activatoi (Data from the National Registry of Myocardial Infarction)

Charles Maynard, PhD, W. Douglas Weaver, MD, Costas Lambrew, MD, Laura J. Bowlby, RN, MBA, William J. Rogers, MD, and R. Michael Rubison, PhD,

for the Participants in the National Registry of Myocardial Infarction*

Very early administration of thrombolytic therapy for acute myocardial infarction (AMI) has significantly re- duced mortality in eligible patients. The purpose of this study was to evaluate factors which influenced the time from symptom onset to hospital presentation and the time from hospital presentation to the onset of throm- bolytic treatment in a lar AMI. This study include cf

e population of patients with 212,990 patients from 904

hospitals that participated in the National Registry of Myocardial Infarction. The median time from symptom onset to hospital presentation for those treated was 1.5 hours versus 2.7 hours for those not receiving throm- bolytic treatment. Older patients and women had in- creased delay times, as did those who arrived at the hospital during daytime hours. Of the 59,802 (28%) patients who received thrombolytic treatment, 23% were treated <30 minutes from admission; 63%, ~60

minutes; and 83%, ~90 minutes. Time to treatment in- creased with age and was longer for women and for patients arriiing between midnight and earl The most important factor associated with I

morning. s orter time

to treatment was the initiation of thrombolytic treatment in the emergency department rather than in the coro- nary care unit (47 vs 73 minutes, p <O.OOOl). Hospi- tal treatment times are much too long, given that quick identification and treatment of eligible patients are of primary importance in reducing mortality from AMI. To shorten these times, thrombolytic treatment should be initiated in the emergency department, and the effec- tiveness of hos ital programs aimed at reducing time to treatment s R ould be subject to continuing quality improvement surveillance.

(Am J Cardiol 1995;76:548-552)

V ery early administration of thrombolytic therapy for acute myocardial infarction (AMI) has significant-

ly reduced mortality in eligible patients.1,2 Barriers to the delivery of thrombolytic therapy in the first hour of symptoms include patient delays in recognizing symp- toms and seeking care, transport time to the hospital, and in-hospital delays in delivering thrombolytic therapy. Several strategies have been employed to attenuate the effects of these barriers. Mass media campaigns have been implemented to reduce patient delays by providing information about recognition of chest pain and the availability of emergency medical services.3,4 To reduce delays in administering thrombolytic therapy, 3 ran-

From the University of Washington Medical Center, Seattle, Wash- ington; Maine Medical Center, Portland, Maine; Genentech, Inc., South San Francisco, California; University of Alabama Medical Cen- ter, Birmingham, Alabama; and ClinTrials Research, Inc., Lexington, Kentucky. The National Registry of Myocardia) Infarction is support- ed by Genentech, Inc., South San Francisco, California. Manuscript received Mav 1, 1995: revised manuscriot received and acceated June 19, 1905.

Address for reprints: Charles Maynard, PhD, MIT1 Coordinating Center, 1910 Fairview Avenue East 205, Seattle, Washington 98102.

*A complete listing of registry hospitals is available from ClinTrials Research Inc., Lexington, Kentucky 40504 (l-800267-5859).

domized trials comparing prehospital and hospital deliv- ery have been conducted.5-7 This study evaluates fac- tors influencing the time from symptom onset to hospi- tal presentation and time from hospital presentation to the initiation of thrombolytic treatment.

METHODS Patient population: In 1990, Genentech, Inc. instituted

the National Registry of Myocardial Infarction (NRMI), a Phase IV observational effort that prospectively col- lects information on the treatment of patients with AMI. At each participating hospital, a registry coordinator completes a l-page form for each patient discharged with the diagnosis of AMI, which is determined by local hos- pital criteria. Data forms are then forwarded to a central data collection facility (ClinTiials Research, Inc., Lex- ington, Kentucky), which manages and analyzes this information. In brief, the registry is used (1) locally, by participating hospitals to assess, and ideally, improve their individual practice patterns; (2) nationally, to mon- itor temporal trends in management practices, with a par- ticular emphasis on thrombolytic therapy; and (3) by Genentech, Inc. to provide safety data for its thrombolytic agent, recombinant tissue plasminogen activator @t-PA) (Activase@; Genentech, Inc., South San Francisco, Cal- ifornia: generic = alteplase, recombinant). A description of data collection procedures has been published.8

548 THE AMERICAN JOURNAL OF CARDIOLOGY@ VOL. 76 SEPTEMBER 15, 1995

This study reports findings from 904 participating hospitals and 212,990 patients with AM1 enrolled in the NRMI from January 1990 to June 1994, and includes the subgroup of patients whose times of symptom onset, pre- sentation, and initiation of thrombolytic treatment with rt-PA were recorded. Although this study includes pa- tients who received all types of thrombolytic agents, only detailed administration times were recorded for patients receiving rt-PA. Individuals whose times from symptom onset to hospital arrival exceeded 24 hours were exclud- ed from this analysis, since these patients may have had poorly described symptoms and/or developed AM1 as the result of hospitalization for other illnesses. In addi- tion, patients who received thrombolytic therapy >3 hours after hospital arrival were excluded, since they probably did not undergo the usual emergency depart- ment triage and treatment process and therefore were not typical patients.

Study variables: In addition to the times of symptom onset, hospital presentation, and treatment, patient age, sex, infarct location, hospital geographic location by re- gion of the United States, hospital bed size, whether treatment was initiated in the emergency department or coronary care unit, and whether the dosage of rt-PA was given over 90 (accelerated or front loaded) or 180 min- utes were collected. Time of chest pain onset was defined as the time when the patient stated that chest pain inten- sified or became prolonged or intolerable, provoking the patient to seek treatment. Time of initial hospital pre- sentation was defined as the hour and minute of initial presentation to the registry hospital, or at the original hospital of presentation if the patient was subsequently transferred to the registry hospital. Approval of the reg- istry data collection process at participating hospitals may have included human subjects review by institu- tional review boards.

Statistical methods: Both mean (&SD) and median (25th and 75th percentiles) times are reported, since the distribution of the time from symptom onset to hospital arrival was not normal. The Kruskal-Wallis test was used to compare medians. Multivariate logistic regression was used to identify predictors of treatment with throm- bolytic therapy within 1 hour of hospital admission. The effects of age, sex, infarct location, time of day, geo- graphic region, place of initiation of treatment (emer- gency department vs coronary care unit), and dosing scheme were considered.

RESULTS Time from symptom onset to hospital presentation:

There were 212,990 patients available for study, and 59,802 (28%) were treated with r&PA. The average age of all patients was 64.9 & 13.4 years and 65% were male. Patients who received thrombolytic therapy were younger (60 vs 68 years of age), more often male (73% vs 61%), and arrived at the hospital sooner than those who did not receive thrombolytic therapy. For patients not receiving thrombolytic therapy, the time from chest pain onset to hospital arrival was longer for patients who arrived from 6 A.M. to 6 P.M. and were admitted to large hospitals in the Mid-Atlantic and Southeast regions

r- ‘ABLE I Median Time (minutes) from Symptom Onset to Hospital irrival for Patients With Acute Myocardial Infarction*

Thrombolysis Not Treated All Treated (no. of (no. of

Jariable [no. of patients) patients) patients)

4dmission time7 12 A.M.-3 A.M. (5,E4) 140 120

(1 1,346) (18,264) 3 A.M.-6 A.M. (4,&) 140 120

(10,131) (16,659) 6 A.M.-9 A.M. (7,92584) 165 125

(14,591) (24,601) 9 A.M.-l 2 P.M. (10::28) 180 135

(22,987) (37,604) 12 P.M.-3 P.M. (10::66) 180 125

(2 1,462) (36,634) 3 P.M.-6 P.M. (9,FZ3) 180 124

(18,834) (3 1,636) 6 P.M.-9 P.M. (7,; 150 120

1) (15,454) (25,536) 9 P.M.-l 2 A.M. (sp7035, 140 120

(14,028) (22,056) Hospital size

1100 beds (2,;5) 150 120

(4,9 151 (8,134) 10 l-200 beds (8,82878) 160 120

(14,710) (25,696) 20 l-300 beds (13512070, 157 120

(28,213) (46,8 14) a300 beds (35:;28) 165 125

(80,759) (13 1,934) Geographic region

Northeast (4,F& 155 120 (10,980) (17,924)

Great takes (10;;92) 158 120 (18,521) (32,131)

Mid-Atlantic (7,9632, 175 135 (16,590) (26,011)

Southeast (6,;7) 180 136 (15,903) (26,054)

South (7,Zl) 163 120 (14,499) (24,393)

Midwest $6) 150 121 (19,974) (31,522)

Southwest (6,:6, 159 120 (12,422) (22,175)

West (7,; 150 120 1) (19,944) (32,780)

*Data from 212,990 patients who participated in the National Registry of Myocordiol Infarction.

tPatients admitted at the very end of a given time group were placed in the following time group [i.e., o patient admitted at 3 A.M. was placed in .he 3 A.M.-6 A.M. group).

(Table I). The median time from symptom onset to hos- pital presentation for those not treated was 2.7 (mean, 4.7 f 5.3) hours compared with 1.5 (mean, 2.4 + 2.8) hours for those who received thrombolytic drug treat- ment (p <O.OOOl).

Older patients and women had increased times from symptom onset to initial hospital presentation (Figure l), as did those patients arriving during daytime hours (Fig- ure 2). The median delay increased from 1.8 hours in patients ~60 years old to 2.5 hours in patients >75 years old (p <O.OOOl). Only 26% of patients presented within

CORONARY ARTERY DISEASE/FACTORS INFLUENCING TIME TO TREATMENT 549

TABLE II Time (minutes) from Hospital Presentation to Treatment for Patients With Acute Myocardial Infarction*

Number 25, 75th Variable of Patients Median Percentile Mean * SD

Sex Men 43,244 49 31, 75 57.4 * 35.8 Women 15,969 55 35, 83 63.2 i 37.2

Hospital size ~100 beds 2,435 50 31, 78 59.2 + 37.5 101-200 beds 8,278 A8 30, 75 57.0 + 36.6 201-300 beds 13,270 50 32, 75 57.8 zt 35.2 >300 beds 35,728 50 33, 78 59.8 f 36.5

Geographic region Northeast A,557 55 35, 83 62.6 f: 37.3 Great lakes 10,792 52 35, 80 60.7 AZ 36.2 Mid-Atlantic 7,422 51 34, 78 59.8 f 35.6 Southeast 6,747 50 30, 77 58.3 * 36.8 South 7,871 50 33, 79 60.1 * 37.6 Midwest 7,896 50 33, 75 58.7 * 36.0 Southwest 6,526 48 30, 75 56.9 zt 35.7 West 7,991 45 30, 70 55.1 * 34.9

Year Jan.-June 1990 698 59 40, 90 67.3 i: 37.3 July-Dec. 1990 3,693 56 36, 85 63.8 + 37.1 Jan.-June 1991 6,223 53 35, 80 61.5 * 36.7 July-Dec. 1991 5,648 55 35, 80 62.3 A 37.3 Jan.-June 1992 7,247 53 35, 80 61.2 f 36.5 July-Dec. 1992 7,099 52 35, 80 60.6 i 36.4 Jan.-June 1993 9,686 A9 32, 75 58.2 i 36.1 July-Dec. 1993 11,323 46 30, 72 55.8 f 35.5

Place of treatment Emergency 47,778 47 30, 70 54.7 * 33.1

department Coronary care unit 9,444 73 45, 1,101 79.2 f 43.3

*Data from 59,802 patients treated with rt-PA in the National Registry of Myocordial Infarction. Due to missing data, the number of patients for o given variable is ~59,802 in some cases.

1 hour of symptom onset. Sex differences in the time to hospital presentation were least apparent in the youngest age group. Delay time was 5 minutes longer for all patients admitted to hospitals with more than 300 beds and was 15 minutes longer for the Mid-Atlantic and Southeast regions of the United States.

6 T

0’

Men Women Men Women Men Women <60 60-75 >75

FIGURE 1. Median time (with 25th and 75th percentiles) from symptom onset to hospital presentation by age and sex. Data from 2 12,990 patients from 904 hospitals that participated in the National Registry of Myocardial Infarction.

TABLE Ill Predictors of Time from Hospital Presentation to Thrombolytic Treatment 51 Hour from Hospital Arrival in Patients With Acute Myocardial Infarction

Factor Odds Ratio 95% Cl

Treatment in the ED 3.08 2.9A, 3.23 Male sex 1.26 1.21, 1.31 Accelerated 90-minute infusion 1.26 1.24, 1.28 Treatment from 6 A.M. to 6 P.M. 1.26 1.22, 1.31 Western United States region 1.24 1.17, 1.30 Advanced age* 0.86 0.83, 0.88 Anterior infarct location 0.90 0.87, 0.94

*Age coded as (1) ~60, (2) 61-74, and (3) 275 years of age. Cl = confidence interval; ED = emergency department.

Time from hospital presentation to treatment: Of the 59,802 patients who received rt-PA, 23% received treat- ment ~30 minutes from hospital presentation; 63%, ~60 minutes; and 83%, ~90 minutes. Only 10% of patients were treated ~1 hour from symptom onset. Time to treat- ment increased with age and was longer for women (Fig- ure 3). Time to treatment also was longer during the mid- night and early morning hours (Figure 2), but did not differ appreciably according to bed size of hospital. Time to treatment was shorter in the West region and decreased noticeably from 1990 through 1993 (Table II). The most important factor associated with reduction in time to treatment was initiation of thrombolytic drug treatment in the emergency department rather than the coronary care unit (47 vs 73 minutes, p <O.OOOl); 83% received thrombolytic therapy in the emergency department.

There were 70 hospitals that had participated in the NRMI since 1991 and had enrolled >lO patients in each 6-month interval from January 1991 to December 1993. The median time from hospital presentation to treatment for the fist enrollment interval was compared with that for the last interval. There was a 9-minute reduction in the median time to treatment (52.5 vs 43.5 minutes); 52 hospitals showed a reduction in median time, 17 had an increase, and 1 hospital exhibited no change (p ~0.001).

40 t

20 I

0

12-3AM 3-6AM 6-9AM 9-12PM 125PM 34PM 6-9PM 9MN

Time of Day

FIGURE 2. Median time from sym tom onset to hospital pre- sentation (N = 212,990) and median time from hospital pre- sentation to treatment (n = 59,802) by time of day. Data from patients from 904 hos Registry of Myocardia P

itals that participated in the National Infarction. MN = midnight.

550 THE AMERICAN JOURNAL OF CARDIOLOGY” VOL. 76 SEPTEMBER 15, 1995

Multivariate analysis: Multivariate logistic regression analysis of the predictors of time to treatment was per- formed in 56,124 (94% of patients receiving rt-PA) patients with complete information (Table III). In order of importance, treatment in the emergency department, male sex, the accelerated 90-minute infusion, treatment during daytime hours, West region, younger age, and nonanterior AM1 location were all associated with treat- ment within 1 hour, although the first independent vari- able was by far the most powerful predictor. Patients treated in the emergency department were 3 times (odds ratio = 3.08, 95% confidence interval = 2.94 to 3.23) more likely to be treated within 1 hour than were those treated in the coronary care unit.

DISCUSSION Despite considerable lay press and medical education

about the benefits of early treatment with thrombolytic therapy, the findings from this large registry demon- strated that >50% of patients delayed >2 hours before seeking medical care; remarkably, 25% of patients delayed >4.7 hours. In this evaluation of patient delays, older age, female sex, and symptom onset during day- time hours were each related to longer delay times. These results were skewed by a large number of individuals who waited very long times before seeking medical care.

In contrast to patient-related delays, hospital person- nel have more control over the time from hospital pre- sentation to initiation of thrombolytic treatment. Despite recognition that treatment times are too long, only 23% of patients were treated ~30 minutes from hospital arrival, and 63% ~1 hour from arrival. An encouraging trend was the 9-minute reduction in time to treatment in a core group of hospitals participating over the 3-year period. Whether this trend reflects a general increase in awareness among clinicians, enhancements to or increased availability of diagnostic equipment, or sim- ply the continuous effect of participation in the registry is unclear.

‘, y------ (I) 6o g 50

; 40 I/

30 t 20

10

0 1 I Men Women Men Women Men Women

~60 60-75 275

FIGURE 3. Median time (with 25th and 75th percentiles) from hospital presentation to treatment by age and sex. Data from 59,802 Nationa P

atients from 904 hospitals that participated in the Registry of Myocardial Infarction.

Rapid treatment is possible and short treatment times have been routinely achieved by hospitals participating in 3 prehospital trials, as well as in hospital-based stud- ies aimed at examining this problem.5-7,9-” Recent stud- ies evaluating the sources of these delays showed that inordinate amounts of time were consumed in obtaining an electrocardiogram by technicians not present in the emergency department, preliminary bedside cardiologist consultations, unnecessary laboratory testing such as chest roentgenograms, and preparation of the infusion in an off-site pharmacy.

In this study, delaying the initiation of thrombolytic therapy to admission to the coronary care unit was asso- ciated with substantially longer treatment times. This finding was true in both univariate and multivariate analyses and has been previously reported.l* It is possi- ble, however, that unmeasured factors (eg, severity of ill- ness, cardiogenic shock) could explain the disparity in treatment times. The only situation in which adminis- tration of thrombolytic therapy in the coronary care unit could be recommended is when the patient is admitted directly to the coronary care unit. Such an approach may be possible for those patients initially evaluated by para- medics who have obtained and transmitted an electro- cardiogram prior to hospital arrival. However, >50% of patients currently do not call 911 and present initially to the emergency department for evaluation of chest pain.3

Study limitations: Several limitations of the NRMI should be acknowledged. First, voluntary participation in the registry resulted in a greater proportion of larger hospitals than expected by chance; >45% of hospitals in the registry had >300 beds, whereas among nonregistry hospitals, only 16% had ~300 beds.8 Second, data forms were not independently audited, and third, the informa- tion collected was modest and therefore limited multi- variate analyses. Fourth, given the large numbers of patients and minimal patient identifiers, there was poten- tial for duplicate entry of patients who transferred from 1 registry hospital to another. Finally, since treatment times were reported only for patients receiving &PA, these results may not be generalized to patients receiv- ing other thrombolytic agents.

The major strength of this report was that findings were generated from large numbers of patients from over 900 hospitals (12% of all hospitals in the United States) representing all parts of the country. In comparison, a recent report concerning prehospital delays and time to treatment was based on 3,868 patients from 8 cities in the United States; >80% of these patients were from 1 metropolitan area. l3 In comparison with other similar studies of prehospital delay and time to treatment, the current one is probably most representative of what is occurring in the United States today.14

Acknowledgment: Drs. Weaver, Lambrew, and Rogers serve as advisors to Genentech, Inc. for the National Reg- istry of Myocardial Infarction.

1. Gtuppo Italiano per lo Studio della Streptochinase nell’lnfarcto miocardio (GISSI). Effectiveness of intravenous thrombolytic treatment in acute myocardial infarction. hncet 1986;1:397402.

CORONARY ARTERY DISEASE/FACTORS INFLUENCING TIME TO TREATMENT 551

2. ISIS-2 (Second International Study of Infarct Survival Collaborative Group). Randomized trial of intravenous streptokinase, oral aspirin, both or neither among 17,187 cases of suspected acute myocardial infarction: ISIS-2. Lancer 1988;2: 349-360. 3. Ho MT, Eisenberg MS, Litwin PE, Schaeffer SM, Damon SK. Delay between onset of chest pain and seeking medical care: the effect of public education. Ann Emerg Med 1989;18:727-731, 4. Herlitz J, Hartford M, Blohm MA, Karlson BW, Ekstrom L, Risenfors M, Wen- nerblom B, Leupker RV, Holmberg S. Effect of a media campaign on delay times and ambulance use in suspected acute myocardial infarction. Am J Cardioll989;64: 9&93. 5. Weaver WD, Cerqueira M, Hallstrom AP, Litwin PE, Martin JS, Kudenchuk PJ, Eisenberg M. Prehospital-initiated vs hospital-initiated thrombolytic therapy. JAMA 1993;270:1211-1216. 6. The European Myocardial Infarction Project Group. Prehospital thrombolytic therapy in patients with suspected acute myocardial infarction. N Engl JMed 1993; 329:383-389. 7. GREAT Group. Feasibility, safety, and efficacy of domiciliary thrombolysis by general practitioners: Grampian Region Early Anistreplase Trial. Br Med J 1992; 305548-553. 8. Rogers WJ, Bowlby LJ, Chandra NC, French WJ, Gore JM, Lambrew CT, Rubi- son RM, Tiefenbmnn AJ, Weaver WD. Treatment of myocardial infarction in the United States (1990-1993): observations from the National Registry of Myocardial

Infarction. Circulation 1994;90:2103-2114. 9. Moses HW, Bartolozzi JJ, Koester DL, Colliver JA, Taylor GJ, Mike11 FL, Dove JT, Katholi RE, Woodruff RC, Miller BD. Reducing delay in the emergency room in administration of thrombolytic therapy for myocardial infarction associated with ST elevation. Am .I Cardiol 1991;68:251-253, 10. Gonzalez ER, Jones LA, Omato JP, Bleecker GC, Strauss MJ (Virginia Throm- bolytic Study Group). Hospital delays and problems with thrombolytic administra- tion in patients receiving thrombolytic therapy: a multicenter prospective assess- ment. Ann Emerg Med 1992;21:1215-1221. Il. Pell ACH, Miller HC, Robertson CE, Fox KAA. Effect of “fast track” admis- sion for acute myocardial infarction on delay to thrombolysis. Br Med J 1992; 304:83-87. 12. Sharkey SW, Brunette DD, Ruiz E, Hession WT, Wysham DG, Goldenberg IF, Hodges M. An analysis of time delays preceding thrombolysis for acute myo- cardial infarction. JAMA 1989;262:3171-3174. 13. Kereiakes DJ, Weaver WD, Anderson JL, Feldman T, Gibler B, Aufderheide T, Williams DO, Martin LH, Anderson LC, Martin JS, McKendall G, Sherrid M, Greenberg H, Teicbman SL. Time delays in the diagnosis and treatment of acute myocardial infarction: a tale of eight cities report from the Pre-hospital Study Group and the Cincinnati Heart Project. Am Heart J 1990;120:773-780, 14. National Heart Attack Alert Program Coordinating Committee, 60 Minutes to Treatment Working Group. Emergency department: rapid identification and treat- ment of patients with acute myocaxlial infarction.Ann Emerg Med 1994;23:3 1 l-329.

552 THE AMERICAN JOURNAL OF CARDIOLOGY@ VOL. 76 SEPTEMBER 15, 1995