There is a growing appreciation of the complex relationship between socioeconomic status and human health. Socioeconomic deprivation has been associated with a higher incidence of coronary artery disease (CAD) and a poorer outcome in patients with CAD.1^–3 However, the influence of social background on outcome from cardiac surgery remains controversial. Higher 3-year mortality rates have been reported after coronary artery bypass grafting (CABG) among disadvantaged patients.4 On the other hand, it was recently found that uninsured patients had a better long-term survival after CABG than insured patients.5 To add to the controversy, socioeconomic deprivation was reported to increase 30-day mortality rate after CABG in one study6 while it had no significant effect in another.7 The influence of socioeconomic status on outcome from valve replacement is not known. The aim of the present study was to evaluate the possible impact of socioeconomic status on mortality after first-time single aortic or mitral valve replacement in the United Kingdom (UK). Furthermore, we aimed to evaluate any differences in outcome between men and women.

METHODS

Data from the UK Heart Valve Registry, which is a national government-funded database set up in 1986 and based at the Hammersmith Hospital, London, to prospectively collect data from all UK cardiac surgical centres, were used. The patients are followed up for the occurrence of death via the national agencies (Office for National Statistics (England and Wales), General Register Office (Scotland), and Central Services Agency (Northern Ireland)) responsible for registering all deaths of UK residents and nationals. These agencies return to the Registry a copy of the patient’s death certificate, which includes date, place and certified cause of death and identifies patients who underwent a post-mortem examination. Thus, the Registry is provided with precise data from implant to death for all patients registered on the database.

Between 1 January 1986 and 31 December 2001, consecutive patients who underwent first-time single aortic or mitral valve replacement were registered on the UK Heart Valve Registry. Only patients with a known and recorded postcode were included in this study. Analysed data included patient age at operation, gender, valve position, type of valve implant, follow-up time, and date and cause of death. The Carstairs deprivation score, a system based on a composite of socio-economic parameters including social class, overcrowding, car ownership, and unemployment,8 was used to stratify cases by level of social deprivation. The score was calculated from the 1991 Census data for the UK (Office of National Statistics) and is based on enumeration districts. The smallest unit of a local authority’s geographic sub-unit hierarchy is the postcode, which is recorded for each patient’s residence. On average 11 postcodes are grouped to form an enumeration district which comprises about 170 households or 400 people. Nationally the score is divided into five levels, with level one being the least and level five the most deprived status.8

Statistical analysis

Kaplan–Meier actuarial outcome curves were constructed to determine 30-day mortality and long-term survival rate. Cox logistic regression was conducted to identify significant risk factors for mortality within 30 days and risk factors with p<0.05 considered to be significant. Multivariate analysis was conducted to identify factors determining long-term mortality. Student t test or χ2 test was used for comparison between groups. Data from the UK Heart Valve Registry database were analysed using Stata 6 Release Software (Stata Corporation, Texas, USA).

RESULTS

Patient population and characteristics

Between 1 January 1986 and 31 December 2001, 51 844 consecutive patients underwent first-time single valve replacement in the UK and were registered on the Heart Valve Registry database.9 The study was 98.3% complete with 861 patients lost to follow-up giving a total number of 268 830 patient follow-up years. The patient population comprised 29 089 men and 22 755 women. Mean age at operation was 64.6 years (SD 11.8) with a range of 18–95 years. Aortic valve replacement accounted for 72% (n = 37 079) of implants compared with 28% (n = 14 765) mitral valve replacement. More men (n = 23 565, 81%) than women (n = 13 514, 59%) had aortic valve replacement (p<0.0001) and conversely more women (n = 9241, 41%) than men (n = 5524, 19%) had mitral valve replacement (p<0.0001). Mechanical valves remained the valves of choice with 31 663 (61%) being implanted versus 20 181 (39%) bioprostheses, with no difference between men and women.

Overall survival and socioeconomic groups

Overall survival at 15 years was 40% with no significant difference between men and women (40% and 39%, respectively). Patients who underwent aortic valve replacement demonstrated a higher long-term survival than those with mitral valve replacement (42% versus 36%, respectively, p<0.0001). Furthermore, patients who underwent valve replacement with a mechanical valve had a significantly higher survival rate than patients with a bioprosthesis (48% versus 27%, p<0.0001).

Long-term survival rate was significantly higher in the least deprived socioeconomic level (Group 1) than in the two most deprived levels (table 1). There was no difference in survival/mortality rate between the least deprived level and the two intermediate levels whereas survival rate was significantly better in the combined intermediate levels than in the combined two most deprived levels (p<0.0001). For reasons of simplicity we based all further calculations on three socioeconomic groups: Group 1 – the least deprived level; Group 2 – the two intermediate levels; and Group 3 – the two most deprived levels. Survival rates of Groups 1 and 3 continued to diverge throughout the observation period (fig 1 and table 2). The difference in outcome between Groups 1 and 3 was mainly due to a lower long-term survival rate (by 18%) amongst women in the most deprived group, that is 35.9% (95% CI: 32.4 to 39.4) versus 43.7% (95% CI: 38.1 to 49.2, p<0.004) in the least deprived group at 15 years’ follow-up. Among men, long-term survival rate in the most deprived group of 39.5% (95% CI: 36.4 to 42.5) was 7% lower than in the least deprived group (42.7%, 95% CI: 37.7 to 47.7, p<0.005). There was no significant difference in overall survival between the least deprived group and the intermediate group or between the intermediate group and the most deprived group with respect to either women or men (data not shown). A poorer long-term outcome was related to socioeconomic deprivation whether or not patients were younger or older than 70 years, whether or not patients had mechanical or bioprosthetic valves, and whether or not patients had aortic or mitral valve replacement (table 3). Although survival rate at 30 days (94–95% across socioeconomic groups) and at 1 year (89–90%) did not differ between any of the socioeconomic groups, it differed significantly between Groups 1 and 3 from 5 years onwards and between Groups 1 and 2 at 15 years’ follow-up (table 4).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Overall survival after mitral or aortic valve replacement according to level of socioeconomic status (Group 1 the least, Group 3 the most deprived background). Hazard ratio: 1.10, p<0.0001.

Cardiac versus non-cardiac deaths

The ratio of cardiac to non-cardiac-related deaths in Group 1 (2638/3961), Group 2 (5452/8620), and Group 3 (5631/7600) did not differ: Group 1 vs Group 2 (p = 0.9); Group 1 vs Group 3 (p = 0.34); and Group 2 vs Group 3 (p = 0.18).

Multivariate analysis

In a multivariate analysis of variables associated with mortality after valve replacement the following hazard ratios and 95% confidence intervals were found: biological valve 1.72 (1.66 to 1.78), p <0.0001; mitral position 1.45 (1.40 to 1.50), p <0.0001; female gender 1.10 (1.06 to 1.14), p<0.0001; low socioeconomic status 1.08 (1.03 to 1.13), p = 0.001.

DISCUSSION

We found that socioeconomic status had no impact on survival rate up to 1 year after primary aortic or mitral valve replacement. After 5 years, however, the survival rate of the most deprived group (Group 3) decreased in relation to that of the least deprived group (Group 1) with a continuous widening of the survival curves until the end of the observation period (fig 1). This difference in survival was most pronounced with respect to women. In the whole population, our study suggests a graded relationship between socioeconomic levels and long-term survival after valve replacement. This relationship was demonstrated irrespective of age, mechanical/bioprosthetic valve, or site of valve replacement. Furthermore, low socioeconomic status was independently associated with a poor long-term outcome.

The UK Heart Valve Registry does not allow for a differentiation between patients with and without coexisting CAD, or whether or not patients with CAD had combined valve replacement and CABG. This could influence outcome of these patients.10 11 Thus patients with concomitant CAD are more likely to be older and to have a history of myocardial infarction than patients with isolated valve disease.12 Furthermore, patients undergoing the combined operative procedure are likely to have prolonged global cardiac ischaemia time, which could affect operative survival. The influence of co-existing CAD has been reported to have a negative impact on survival rate if the patients are not revascularised in relation to valve replacement.12 13 According to The Euro Heart Survey on Valvular Heart Disease, fewer than a quarter of patients undergoing single aortic or mitral valve replacement have additional CABG.14 As ischaemic heart disease generally has a higher prevalence in a disadvantaged population, a higher rate of coexisting CAD would be expected among valve disease patients from a disadvantaged background, and therefore a higher operative mortality rate in this group.1^–3 Either this was not the case, or coexisting CAD did not play a major role for the operative survival after valve replacement in our cohort, irrespective of socioeconomic level.

Living in a poor community has been reported a mortality risk factor 36 months after bypass grafting.4 There are several possible reasons for a poorer long-term outcome for disadvantaged patients after major cardiac surgery. Low socioeconomic status has been associated with lower rates of exercise and higher rates of smoking.15 16 Also higher lipid levels, a higher prevalence of obesity and diabetes, and higher blood pressure levels amongst the socially most deprived may all contribute to a higher long-term mortality rate.15^–17 Furthermore, several studies have shown that economically disadvantaged subjects are less likely to benefit from evidence-based advances in medical care or undergo invasive cardiac evaluation and revascularisation than their less deprived counterparts.2 18 19 It has been suggested that poor people (not covered by a comprehensive care programme) may not be able to afford prescription drugs.16 Furthermore, diseases other than coronary artery disease (eg cancer and infections) have been reported more frequently among patients from socially deprived areas and may contribute to a poorer long-term outcome.20

Limitations

In order to maximise the rate of reporting, the database requested a minimum of clinical information. We were therefore not able to measure the effect of socioeconomic status in relation to other well-known risk factors (eg poor left ventricular function, presence of CAD, and other important comorbidity) after valve replacement. The strength of the study is, however, the total number of patients in the database and the fact that the database covers all of the UK. This allows for comparison of our results with those from other countries, providing a global perspective. Although we used a small area index as a proxy to individual socioeconomic status in order to stratify patients into socioeconomic levels, the Carstairs score has been widely used and validated for this purpose in the UK.8 Our socioeconomic grading is based on the address of the patients at the time of the operation. We had no means to assess the possible effect of a change of address and thereby a change of socioeconomic status on mortality.21 Despite this inherent weakness our study suggests a sustained effect of the socioeconomic level achieved at the time of the operation.

Conclusions

The impact of disadvantaged social background on mortality after valve replacement became evident from 5 years onwards after the operation, especially among women. This precludes any important effect of operative trauma in people from a cross-section of socioeconomic backgrounds. We suggest that an increased number of risk factors among the socially deprived, so important in relation to health and survival in the general population, play a similar role for survival of disadvantaged patients with artificial heart valves. An invigorated approach to the management of risk factors and improved access to medical care may improve long-term survival rates after valve replacement in these patients.