Intervention Program to Improve Secondary Prevention of Myocardial Infarction. Results of the PRESENTE (Early Secondary Prevention) Study

INTRODUCTION

The 4S,¹ CARE,² and LIPID³ studies have shown that lowering cholesterol levels by prescribing statins for patients with chronic coronary heart disease reduces rates of all-cause mortality, coronary heart disease mortality and cardiac mortality. Statin therapy also decreases reinfarction rates and the need for heart surgery.

Cardiology Societies believe sufficient evidence exists to prove the relationship between levels of total cholesterol, low density lipoprotein (LDL) cholesterol and cardiovascular risk and recommend maximum doses for medication.^4-6 Beta-blockers,^7-9 angiotensin converting enzyme inhibitors (ACE inhibitors),¹⁰ and antiplatelet drugs^11,12 have also proved beneficial in secondary prevention.

However, a number of patients who should benefit from these findings and recommendations do not actually do so. The PREVESE II study¹³ showed that only 29% of patients with myocardial infarction were prescribed statins at discharge, 46% were prescribed ACE inhibitors, and 45% beta-blockers. However, 88% of patients were prescribed antiplatelet drugs, which is a figure much closer to recommendations. The EUROASPIRE II study¹⁴ reported similar patterns in several European countries, as did the US National Registry of Myocardial Infarction¹⁵.

The causes of this appear to be physicians' prescriptions and patients' compliance with treatment. Early prescription would improve the situation. In France, the PREVENIR study¹⁶ showed that 95% of patients prescribed lipid-lowering medication at hospital discharge continued to comply with treatment at 6 months. Statin therapy is the recommended secondary prevention treatment least often prescribed at discharge. This may be due to the spontaneous decrease in cholesterol levels during the acute phase of myocardial infarction and the lack of studies that recommend statins for patients with acute coronary syndrome. However, starting statin therapy in the acute phase of coronary heart disease^18-20 has been demonstrated to have an early beneficial effect on endothelial dysfunction¹⁷ with an almost complete absence of side effects.

Secondary prevention using different, more or less complex interventions can produce significant improvements as recent research has shown.^21-23 The objective of our study was to determine whether heightening the awareness of attending physicians followed by two simple interventions (obtaining lipid levels within 24 hours of admission and providing patients and their families with educational information before discharge) would improve secondary prevention at discharge and at 6-month follow-up.

We also analyzed patient satisfaction with the educational program and recorded adverse cardiovascular events occurring during the 6-month follow-up.

PATIENTS AND METHODS

Almost all hospitals in Spain were invited to participate in this study. Participating hospitals had a cardiology service or section and a coronary or intensive care unit, were able to determine patients' lipid levels on admission, received a minimum weekly average of 5 patients with myocardial infarction and agreed to introduce an intervention program to provide information and health education. Initially, 120 hospitals agreed to participate but 10 were subsequently excluded because data submitted could not be identified or because no follow-up took place. Patient enrolment began in September 1999 in the belief that each center would reach a maximum of 50 consecutive patients within 3 months.

Sample size was calculated according to the percentage of patients who at the time of hospital discharge were able to start statin therapy, bearing in mind the evolution observed in the PREVESE studies,¹³ in which the proportion grew from 7% to 30%. In this study, we estimated that statin prescription at discharge might approach 50%. With a sample of 2965 patients this would produce a 95% confidence interval ≥1.8%. We calculated a minimum sample size of 3488 patients on the assumption that prescriptions would continue at 50% for 6 months with new enrolments to compensate for a predicted 15% sample erosion.

For each patient we recorded data of hospital affiliation, cardiovascular risk factors (CVRF) prior to acute myocardial infarction (AMI) including lifestyle variables and work activity, family and personal history of coronary heart disease, results of a simple physical examination (weight, height, blood pressure and heart rate), lipid levels determined within 24 hours and treatment at discharge. We also noted whether or not patients were involved in an intervention program to provide information and health education and the 6-month follow-up visit.

At follow-up, we recorded adverse events and patients underwent physical examination and lipid analysis. Pharmacologic treatment and patient observations on compliance with treatment were recorded as were motives for abandoning medication and patient evaluation of the intervention program.

The intervention program to provide information and health education was carried out by attending physicians. In individual interviews, patients and family were taught the importance of modifying CVRF, adopting more adequate lifestyle and dietary habits and adhering to prescribed medication regimens. Patients were given printed leaflets containing the same information and small reminder stickers to place strategically around their homes. Patients were informed they would be given 6-month follow-up appointments and that they should undergo lipid analysis. Contact with physicians who would handle the follow-up was limited to the standard practice of each center. This usually consisted of the discharge summary.

Statistical analysis consisted of a descriptive study of baseline data. Categorical variables were expressed as absolute frequencies and percentages and quantitative variables were given as mean and SD. Independent samples were compared using chi-square for proportions and the Student t-test for means. The McNemar test was used to compare related sample proportions and the Student t-test for repeated measures.

For descriptive purposes, we present the percentages that demonstrate patient satisfaction with the intervention program and the occurrence of cardiovascular events.

Control and statistical analysis of data has been carried out by an independent company (LOGITEST) in collaboration with the study's Scientific Committee which supervised the sequence of patient enrolment in the centers and checked the reliability of the data collected.

RESULTS

A total of 4174 patients discharged from the 110 participating hospitals were enrolled in the study. Distribution of patients by region is shown in Figure 1. We excluded 144 patients who were unidentified or lost to follow-up. The final sample was 4030 patients with an average age of 63.7±11.7 years. Women represented 26.9% of the sample and were significantly older than men (68.2±10.5 vs 62.0±11.7 years). Prevalence of CVRF and differences between men and women appear in Table 1.

Fig. 1. Distribution of participating patients by region.

Baseline family history data was available for 76% of the sample. We recorded antecedents in the histories of 24.2% of patients. Prior myocardial infarction was recorded for 12.3% of patients, 12.4% had suffered angina of >3 months duration, 3.4% had undergone percutaneous coronary revascularization and 1.6% surgery. A total of 40.3% were in work (44.8% men, 26.9% women).

Average weight of patients was 76.2±11.7 kg (n=3616, 89% of sample), average height was 166.8±7.9 cm (n=3552, 88% of sample) and body mass index (BMI) was 27.4±4.0 (n=3542, 87.8% of sample).

A BMI >25 was defined as overweight and 74.4% of women and 70% of men met this criterion. Obesity (BMI>30) was found in 34.1% of women and 20% of men, with statistically significant differences.

Systolic blood pressure was 133.2±21.4 mm Hg and diastolic blood pressure was 76.9±13.0 mm Hg (n=3891). Heart rate at rest was 72.6±12.4 beats/min. At discharge, 71.1% of patients had controlled blood pressure (≤140-90 mm Hg).

Cholesterol levels were determined within 24 hours for 96% of the sample (n=3896) and fractions were determined for 82% (n=3480): total cholesterol was 219 (44.9) mg/dL, high density lipoprotein (HDL) cholesterol was 40±11.5 mg/dL, low density lipoprotein (LDL) cholesterol was 154±37.7 mg/dL and triglycerides averaged 156±67.4 mg/dL. Total cholesterol >200 mg/dL was found in 65% of patients and 88% presented low density lipoproteins (LDL) >100 mg/dL. Treatments administered on hospital discharge appear in Table 2. Informative health education interviews were held with 93.2% of patients prior to discharge. These patients also received informative leaflets.

At 6-month follow-up, we received information on 89.9% of the sample. We identified 65 patients who had died (1.6% of those located) and 5.7% refused to attend. We were unable to follow up 420 patients (10.4%) (Figure 2). Table 3 compares baseline data of the 6-month follow-up patients with data of patients lost to follow-up. On average, patients lost to follow-up were significantly older, more of them were women and the percentage of smokers was lower. No significant differences appeared between the groups for patients with diabetes, high blood pressure or average total cholesterol. Patients who died were significantly older and the number of women and patients with diabetes was higher although there were fewer smokers.

Fig. 2. Flowchart of patients in the study and follow-up  

In addition to the deaths recorded, we found 2.2% of patients had suffered myocardial infarction, 1.5% had had a heart attack requiring hospitalization, 6.9% had suffered angina requiring hospitalization and 2.9% had undergone revascularization. In all, 11% had required hospitalization.

Table 4 describes significant beneficial changes in a range of parameters in patients attending 6-month follow-up appointments. The survey of lifestyle habits showed that 86.5% of smokers reported they had given up smoking; 71.4% of patients stated they took more exercise than previously; and 77.1% affirmed they had improved their dietary habits.

Table 2 also compares pharmacologic treatment at 6 months with treatment at discharge (n=3193). We found that compliance with prescriptions had been maintained and that levels of prescription were high for statins (88%) and antiplatelet drugs (93.8%). A slight but significant reduction in administration of ACE inhibitors was found but this was balanced by an increase in angiotensin II receptor antagonists and a significant reduction in nitrates, diuretics and anticoagulants. Levels of prescription of beta-blockers were maintained.

Compliance with medication was controlled via patients' statements and non-compliance was defined as not having taken the prescribed dose on one day a week or on 24 days over the 6 months. Non-compliance with medication was reported by 21.1% of patients taking antiplatelet drugs, 21.7% of those taking antihypertensive medication and 28% of patients ta-king lipid-lowering drugs. Patients stated non-compliance was due to side effects (3.8%), forgetfulness (7.6%), thinking the medication was no longer necessary (2.1%) and other causes (2.9%).

The majority of patients (89.2%) believed the intervention program had persuaded them to adopt a healthier lifestyle; 92.3% thought it improved their understanding of the importance of complying with therapy; and 92.8% considered it had helped them follow their medication regimen more assiduously. In addition, 87.7% said the informative leaflet was easily understood and 87.3% considered it had provided useful information about their illness.

DISCUSSION

Our results represent an improvement on the PREVESE II study¹³ which did not involve any specific intervention. The most striking improvement has been the increase in the prescription of statins, which has risen from 30% to 88%. The difference in lipid values, which were much higher in this study as they were determined within 24 hours of myocardial infarction, may have influenced this increase. However, we cannot confirm this as the cause of increased statin prescription because no control group was included in the study. The increase in prescriptions may be a consequence of heightened awareness on the part of physicians or the evolution of clinical practice. Notwithstanding, these results suggest that interventions as simple as these can improve secondary prevention in selected samples of patients.

In this context, heightening the awareness of participating physicians may have reminded them of the indication criteria of preventative drugs prescribed at discharge. This is borne out by comparing our results with those of non-intervention studies¹³ because we also found a 5% increase in prescription of antiplatelet drugs and a 14% increase in beta-blockers.

If we compare prescriptions at discharge with data from the non-intervention IBERICA 1997 study²⁴ (considered a model in Spain given its high reliability and 4401 patient population representing all cases of AMI in 7 regions) we find higher levels of prescription of antiplatelet drugs (94.1% vs 91.5%), beta-blockers (59.4% vs 44.5%), and ACE inhibitors (47.6% vs 37.6%).

The increase in statin prescriptions at discharge may have been influenced by the determination of lipid levels during the acute phase of infarction, the confirmation that 88% of patients had LDL cholesterol >100 mg/dL, and the steady increase in information about the safety and efficacy of statins in acute coronary syndromes.^17-19 At 6 months, the early use of statins in our study was responsible for significant reductions in average total cholesterol and LDL cholesterol levels, and the values obtained were much closer to those recommended in secondary prevention. However, as we have indicated, the evolution of clinical practice may have played a part in achieving these results. The EUROASPIRE II study,¹⁴ despite its lack of an intervention program, also found a considerable increase in prescriptions by comparison with data from their earlier study.²⁵

Significant increases in the use of aspirin (rising from 68% to 92%), beta-blockers (from 12% to 62%), ACE inhibitors (from 6% to 58%), and statins (from 6% to 86%) were achieved by a similar intervention study based on a smaller patient sample.²³ Post-AMI patients were informed about the illness and CVRF control and attending physicians were advised to start preventative treatment prior to discharge.

Other parameters of clinical examination have also been seen to improve significantly (Table 4) with decreased average values for blood pressure, weight and BMI, as well as the favourable modifications in lipid levels mentioned. Some 75% of patients said they took more physical exercise and followed a healthier diet, although 13.5% of smokers had either not given up or had started smoking again.

Adherence to pharmacologic prescriptions in secondary prevention fundamentally depends on recommendations given at discharge, as shown by the PREVENIR study¹⁶ which confirms our data. Differences in the prescription of statin therapy after myocardial infarction between a group of patients who received treatment at discharge and a similar group who did not (88% vs 30%) were made evident in a recent publication²⁶.

Different intervention programs have improved secondary prevention measures adopted by physicians and patients. In the United Kingdom,²¹ a special program of nurse-led clinics in primary care was responsible for substantial improvements in treatment compliance and CVRF control (blood pressure, lipid levels, physical exercise, diet), and only failed in relation to smoking. Similarly, a personalized system of 3 monthly reminders to physicians increased the use of statins from 47% to 85%, reducing LDL cholesterol values to 41%.²² In our study, we monitored compliance via patient statements and achieved almost 80% success. In addition, some 90% of patients confirmed that the educational program had helped them follow the medication regimen and improve lifestyle habits.

In 15.9% of patients we recorded some form of adverse cardiovascular event during follow-up, with a mortality rate of 1.6%. The Swedish RIKS-HIA registry,²⁰ including >19 000 patients discharged following AMI, reported a 4% one-year mortality rate among patients who received early statin therapy (29% of the total) by comparison with 9.3% mortality among the rest of the population. Our study did not include a control group of non-intervention patients but the mortality rate does seem similar to that reported for statin therapy patients in the Swedish registry.²⁰ However, sample erosion in the follow-up (10%) detracts from the value of our data. In contrast, the Spanish PRIAMHO registry²⁷ of 5242 patients with AMI without any secondary prevention intervention reported 7.6% one-year mortality after discounting patients who died during hospitalization.

A recently published, open, randomized, prospective study²⁸ of intended treatment shows that patients with coronary heart disease whose LDL cholesterol is held at ≤100 mg/dL for 3 years reduce their all-cause risk of death by 43%, risk of death due to cardiac event by 47% and risk of non-fatal infarction by 59% when compared with a similar population receiving standard attention to cholesterol levels.

These data would suggest that early lipid-lowering treatment in secondary prevention may reduce adverse events, although results of clinical trials have yet to be published.^29,30

LIMITATIONS OF THE STUDY

A wide range of hospitals from all over Spain participated in this study with no prior sampling or randomization. The high level of participation was intentional as we wanted to demonstrate that an intervention program to provide information and health education could easily be introduced in numerous hospitals.

Data on some of the parameters (level of physical activity, diet, treatment compliance) have been obtained without using quantitative instruments which would have complicated managing the registry. However, patient statements about these parameters were congruent with data on weight loss and reduced BMIs, and reduced lipid values were consistent with the prescription of statins.

A total of 715 patients (17.7%) could not be located or refused to attend follow-up appointments and the majority of them were women or older patients. This loss of patients during the follow-up may be linked to a lower level of compliance. Consequently, our results can only be applied to those patients who followed the advice given in the program.

The absence of a control group of hospitals clearly limits the value of our conclusions. We have described changes as a consequence of our intervention program by comparison with other studies but we recognize that other factors may have been involved. A further study involving a control group of non-intervention hospitals is needed to confirm these results.

CONCLUSIONS

In spite of the limitations mentioned, our results suggest that a simple intervention program centered on a wide-ranging group of selected patients who have survived an AMI and whose lipid levels have been determined within 24 hours can produce an improvement in secondary prevention and maintain favorable results over the first 6 months.

Our simple program was widely accepted by patients, most of whom believed it helped them maintain lifestyle changes and adhere to medication regimens.

Scientific Committee and Principal Researchers at Each of the Hospitals Participating in the PRESENTE Study

Scientific committee: José A. de Velasco Rami, Juan Cosín Aguilar, Manuel de Oya Otero, and Eduardo de Teresa Galván.

Centers and Researchers

Hospital del Aire: Manuel González Calvo. Hospital General de Albacete: Miguel A. Barba. Hospital de Alcoy: Guillermo Grau. Hospital Arnau de Vilanova: José F. Sotillo Martí. Hospital Arquitecto Marcide: Gonzalo Pía Iglesias. Hospital de Ávila: José Lázaro Fernández. Hospital de la Axarquía: Lorenzo Pérez Benito. Hospital Provincial de Badajoz: Manuel Anguita. Hospital de Basurto: José M. Aguirre Salcedo. Hospital Baza: Guzmán Arquero. Hospital Bellevue: Luis P. Febles Palenzuela. Hospital de Cabueñes: Alberto Batalla. Hospital Universitario de Canarias: Juan Lacalzada García, Antonio Lara Padrón. Hospital Can Misses: Ismael Laplaza Alastruey. Hospital Carlos Haya: Félix Malpartida de Torres. C.H. de Cardiología: Josep Reig Barbel. Hospital General de Castellón: Juan Moreno Rosado. Hospital Central Asturias: Gustavo Iglesias Cubero y Vicente Barriales Álvarez. Hospital Clínico: Clodoaldo Alburqueque. Hospital Clínico San Cecilio: José L. Martí García. Hospital Clínico Universitario: Javier Merino Sesma, Javier Rodríguez Collado y José A. Ferrero Cabedo. Hospital da Costa: Lombín. Hospital Creu Roja: Eduardo Homs Espinach. Hospital de Cruces: Pedro M. Montes. Hospital Divino Valles: Pablo Ruiz Pérez. Hospital Dr. Negrín: Eduardo Caballero Dorta y Efrén Martínez Quintana. Hospital General de Elche: Fernando García de Burgos. Hospital General Elda: Rubio Pérez. Hospital La Fe: Antonio Flores Pedauye y Luis Martínez Dolz. Hospital de Figueres: Francisco J. Monzón. Hospital General de Fuerteventura: Miguel García Navarro. Hospital de Galdakano: Iñaki Lekuona Goya. Hospital García Orcoyen: Maite Beunza Puyol. Hospital Gregorio Marañón: José A. Serrano Sánchez. Hospital General Yagüe: Carlos Alexandre Almeida Fernández. Hospital Granollers: Arturo Herrero. Hospital General Guadalajara: Javier Balaguer Recena. Hospital de Guipúzcoa: De la Cuesta, Telleira. Hospital Comarcal de Hellín: Fernando Benstez de Castro y María E. de Tomás Labat. Hospital Infanta Elena: Antonio Barrios y José L. Gómez. Hospital Insular: Vicente Nieto Largo. Hospital Josep Trueta: Rafael Masía Martorell. Hospital Juan Ramón Jiménez: José M. Sobrino. Hospital Lluis Alcañiz: Julio García del Busto Arregui. Hospital General Mallorca: Antonio Rodríguez Fernández. Hospital Manacor: Bernardo García de la Villa, Izascun Obieta Fresneda y Rosa Lázaro García. Hospital Marina Alta: Segundo Martí Llinares. Hospital Marqués de Valdecilla: José R. de Berrazueta. Hospital de Mataró: Joan Batle Díaz. Hospital Miguel Servet: Adolfo Marquina, Eva Pue, Luis Escosa y Luis Placer. Hospital de la Merced de Osuna: Alicia Cabello González, Antonio González Barrera y José Quero Cuevas. Hospital Policlínico Miramar: Bartolomé Vallbona Adrover. Hospital Morales Meseguer: José A. Ruiz Ros. Hospital de Móstoles: Ruiz Galiana. Hospital General Universitario de Murcia: Juan García Reverte. Hospital de Navarra: Jesús Berjón Reyero. Hospital Nuestra Señora de Aránzazu: José M. Morán Dacal. Hospital Nuestra Señora de la Candelaria: Julio Hernández, Rafael Romero. Hospital Nuestra Señora del Valme: Luis Pastor, Francisco J. Molano. Hospital 12 de Octubre: José Palacios Martínez. Hospital La Paz: Luis Martín Jadraque. Hospital de la Princesa: Enrique González Pérez. Hospital Princesa Sofía: Julián Bayon Fernández y Esteban García Portero. Hospital Príncipe de Asturias: Juan de Dios García-Díaz. Hospital de Poniente: José Luis Zambrano García. Hospital Puerta de Hierro: Ramón Arroyo Espliguero y Jesús Jiménez Mazuelos. Hospital Puerta del Mar: Carlos Piñero Gálvez. Hospital Puerta Real: Enrique Otero Chulián, Marisal Travedra Bosque, Manuel Maestre García y Rafael Hidalgo Urbano. Hospital Punta de Europa: Rafael Hidalgo Urbano. Hospital Rafael Méndez-Lorca: José L. Sánchez Puerta. Hospital Ramón y Cajal: María C. de Pablo y Zarzosa. Hospital Reina Sofía: Federico Vallés. Hospital del Río Ortega: Miguel A. Cobos Gil. Hospital Sagrado Corazón: Josefa Rafel Rivera. Hospital de Sagunto: Ramón Moragrega. Hospital San Agustín: Víctor M. Rodríguez Blanco. Hospital San Eloy: Javier Andrés Novales. Hospital San Francisco de Borja: Salvador Bellver Bellver. Hospital de San Juan: Vicente Bertomeu Martínez. Hospital San Pedro de Alcántara: Jesús Montero. Hospital Comarcal de Sant Boi: César Romero Menor. Hospital Santa María: Jacint Cabau Rubies. Hospital Santa María Madre: Miguel Pérez de Mon Romero. Hospital Santa María Rosell: Juan Ortega Bernal. Hospital Santa Tecla: Juan Carlos Somero Giménez. Hospital Santiago Apóstol: Santiago Cantabrana. Hospital General Segovia: Carlos Junquera Planas. Hospital Comarcal de Valdeorras: Josep Masferrer Serra. Hospital General Universitario Valencia: José A. Velasco Rami. Hospital Clínico de Valladolid: Joaquín J. Alonso. Hospital Comarcal Vega Baja: José M. Depeda Rodrigo. Hospital Virgen de la Arrixaca: Mariano Valdés Chávarri. Hospital Virgen Macarena: Juan Navarrete Reinoso. Hospital Virgen de la Luz: Sonia Rodríguez Díez. Hospital Virgen del Valle: Luis Martos Leal. Hospital Virgen de la Victoria: Ángela Montijano Cabrera. Hospital Xeral: Juan Vidán Martínez.

*A list of the researchers involved in PRESENTE appears at the end of the article. This studyl support through a grant from Bristol Myers Squibb, Spain.

---

Correspondence: Dr. J.A. de Velasco.
Servicio de Cardiología. Hospital General Universitario.
Avda. Tres Cruces, s/n. 46014 Valencia. España.
E-mail: [e-mail]