gms | German Medical Science

Background: Cardiovascular events, myocardial infarction (MI) in particular, carry significant burden of disease coupled with high mortality rates. Disease management programs (DMPs) have been implemented to reduce MI-related morbidity and mortality and improve secondary prevention. While differences in MI-related survival rates between male and female patients are documented, little is known about complex patterns of interaction between risk factors and gender-specific risk.

Aim: To assess interactions between risk factors for first-time MI and test for gender-specific differences in risk factors for first-time MI in patients suffering from coronary heart disease (CHD), who participate in the CHD-DMP in the region of North Rhine.

Methods: We performed conditional inference survival tree analysis. This classification method allows identifying the optimal threshold of predictor variables through univariate splits in a recursive fashion. Analyses were run in a subsample of CHD DMP-patients, who had not suffered of MI upon inscription in the program and presented with complete data. Time until occurrence of first MI during DMP participation (in years) was used as dependent variable and known risk factors for MI which were routinely collected in the CHD-DMP such as age, age of onset and duration of CHD, blood pressure, smoking behaviour, cholesterol levels and comorbid diseases were included as covariates. First, we tested for gender differences in MI survival rates (n=21891, 65.8% male, mean (age)=76.2 years, SD (age)=8.8). Subsequently, we performed conditional inference survival tree analyses on the whole sample as well as for each gender separately.

Results: Our preliminary results indicate significant gender differences in survival rates for MI (chi²=19.5, p < 0.001). Using data from the whole sample, we identified five subgroups defined by different risk factors profiles.

We detected interactions between the following risk factors: duration of illness, total cholestrol levels as well as the presence of comorbid disorders, all p < 0.001. Of all examined variables duration of illness was identified as the key discriminator. Subsequent splits were obtained at duration of illness of > 12 years, total cholestrol levels of > 165 mg/dl and the presence of comorbid disorders. Gender-specific analyses yielded four and five subgroups for females and males, respectively. Gender-specific analyses confirmed the importance of the aformentioned risk factors. Yet, in female patients different splits significantly differentiated risk profiles of subgroups: duration of illness at > 5 years, total cholestrol levels at > 211 mg/dl, and presence of comorbid disorders, all p < 0.03.

Discussion and practical implication: Our preliminary findings shed light on risk factor profiles for first-time MI in male and female CHD patients, who take part in CHD-DMP. More specifically, we show that similar risk factors are implicated in MI between genders, albeit to different extents. Using routinely acquired DMP data, our findings highlight the importance of assessing duration of CHD illness, cholesterol levels and comorbid disorders in order to foster the prevention of a secondary non-fatal MI.