Assessment of health-related quality of life in Thai patients after heart surgery

In the heart surgery project, 7,863 children and adult Thai patients living across the country were chosen without payment for cardiac surgery, i.e., the valve replacement/repair, operation of atrial and/or ventricular septa, or coronary artery bypass graft (CABG), as shown in Figure 1. Patients were included in this study if they were 15 years old or more and underwent the heart surgery during the previous one year, thus leaving 5,255 patients in the sample pool. Half of the patients were randomly selected by a systematic sampling technique (i.e., patients with even hospital numbers) and invited via mail to participate in the study and provide their telephone numbers. Only 901 patients were willing to take part in the study and half of them were further systemically sampled for a telephone interview. After excluding patients for various reasons, only 386 patients completed the study. All of them provided written informed consent to participate in this study.

Figure 1

Two generic instruments were used because, unlike disease specific tools, they are more flexible in terms of patient population and interventions. The first one was the health profile measure “12-item Short Form, version 2” (SF-12v2; Quality Metric Inc, Lincoln, RI, USA). This health survey, which is the short form of the widely used SF-36, is a brief and reliable measure of overall health status [9]. The Thai version of SF-12v2 was used with the permission from the Quality Metric Inc. It measures eight health domains: physical functioning (PF), role limitations because of physical health (role-physical: RP), bodily pain (BP), general health perceptions (GH), vitality (VT), social functioning (SF), role limitations because of emotional problems (role-emotional: RE) and mental health (MH). To complete the tool, the recall period for the SF-12v2 was within the past four weeks. The SF-12v2 responses were scored from 0 to 100 on each of the domains, together with the Physical Component Summary (PCS) scale and Mental Component Summary (MCS) scale. Higher scores indicated a better HRQoL. To simplify comparisons with the general population, the norm-based scoring was used, i.e., scores being linearly transformed to a scale with the mean of 50 and standard deviation (SD) of 10 for the general US population [10]. In this study, Cronbach’s alpha values for PCS and MCS were 0.73 and 0.77, respectively. Values >0.7 indicate that the SF-12v2 has an acceptable internal consistency reliability [11].

The second instrument was a measure of HU, i.e., EuroQoL (EQ-5D) and Visual Analog Scale (EQ-VAS). The EQ-5D consists of five attributes: mobility, self-care, usual activity, pain/discomfort, and anxiety/ depression [12]. Each attribute has three levels: no problem, some problems, and major problems. Its Thai version was obtained from the EuroQoL Group and the Thai value set was employed to calculate EQ-5D utility scores [13]. A resulting EQ-5D score is usually between −0.45 and 1.00, where 1.00 and 0 represent perfect health and death, respectively. Higher EQ-5D scores indicate better health, but negative values signify the states worse than death. In this study, the Cronbach’s alpha value of the EQ-5D was 0.61, which is considered moderately reliable. The second part of EQ-5D is a visual analog scale (EQ-VAS) that asks respondents to rate their current health states with scores starting from 0 (worst imaginable health) to 100 (best imaginable health). EQ-VAS scores were obtained by dividing the number rated on the scale by 100.

All 386 patients with postoperative heart surgery were interviewed on the phone by trained research assistants using the two generic measures. The telephone interview was conducted for approximately half an hour for each patient in the academic office at the Ramathibodi Hospital. The data gathered were entered into Microsoft Excel and SPSS version 17.0 (SPSS Inc., Chicago, IL, USA) for analysis.

The characteristics of patients were analyzed using descriptive statistics, e.g. percentages, means, and standard deviations (SD). A one-sample t test was used to test the differences between the eight domains of SF-12v2 of the Thai heart surgery sample and those of the U.S. general population, whereas a Chi-squared test was to determine the differences between EQ-5D scores of the heart surgery sample and those of the Thai general population. A paired t test was also conducted to test the difference in EQ-VAS before (patients being asked to recall their health prior to the operation) and after the heart surgery. Moreover, an independent t test was performed, as part of a subgroup analysis based on age cut off by median, sex, work status (employed or unemployed), and presence of commodity (yes or no), to compare the differences for EQ-5D, EQ-VAS, PCS, and MCS. To determine the relationships of EQ-5D and EQ-VAS with eight SF-12v2 domains plus two summary scores, a multiple linear regression analysis using the ‘enter’ method was also performed. A difference with P < 0.05 was considered significant.

Patient characteristics are summarized in Table 1. Of 386 patients, with a mean age was 50.4 ± 13.7 years and nearly half were male (49.5%), which is comparable with that of the entire cohort excluding children (n = 5,225; mean age = 50.3 ± 15.2 years, 50.1% male). Slightly more than half were employed with low monthly wages. As shown in Table 2, high scores for both health profile and health utility measures were revealed in this patient group, as evidenced by the means or medians. Additionally, most SF-12v2 domain scores except ‘General Health’, EQ-5D, and EQ-VAS had high ceiling effects, i.e. being greater than the limit of 15% [14]. In Table 3, compared with general Americans [9], the Thai patients had higher VT, SF, MH and MCS scores, but lower RP, BP, GH, and PCS scores (all P < 0.01 except BP with P < 0.05). No statistical difference in domain scores for PF or RE was found in this group of patients.

Table 1

Patients’ characteristics (n = 386)

Table 2

Descriptive statistics of SF-12v2, EQ-5D, and EQ-VAS

Table 3

Comparison of 8 domains and 2 summary scores of SF-12v2 between the Thai sample and the US general population [ref. 9]

Thai patients after heart surgery possessed a higher EQ-5D score (0.81 in Table 2) than that of Thai patients with ischemic heart diseases (0.75) [15] and with heart failure (0.73) [16], but lower than the Thai general population (0.84) [13] (all P < 0.01; data not shown in the Table). Nevertheless, higher percentages of the patients reported ‘no problems’ in every domain of EQ-5D except for ‘Usual activities’ compared with those of Thais in general (Table 4). For EQ-VAS, the patients reported a great deal of improvement in their health after heart surgery, i.e. from 0.54 to 0.84 (P < 0.01; data not shown in the Table). The EQ-VAS of Thai patients after heart surgery (0.84 in Table 2) was higher than that of the Thai general population (0.77) [13] (P < 0.01). However, it should be noted that the mean age of the general Thai population (44.6 years) and the percentage of males (45.7%) [13] were lower than those of the Thai patients after heart surgery (mean age = 50.4 ± 13.7 and male = 49.5%). Generally, younger male people tended to have a higher HRQoL than older women. Thus, the HU values of the Thai patients measured by the EQ-5D and the EQ-VAS were overall comparable with those of the Thai general population.

Table 4

Comparison of EQ-5D domains between Thai patients after the heart surgery and the Thai general population [ref. 13]

As for the subgroup analysis in Table 5, younger patients (≤ 52 years old) had higher EQ-VAS and PCS scores than older patients (both P < 0.01). Female patients had lower EQ-5D and PCS scores than their male counterparts (both P < 0.01). In addition, patients with employment or no comorbidity tended to have higher HP and HU scores (all P < 0.05). For the relationships between HP and HU (Table 6), the multiple regression models showed that EQ-5D was significantly associated with all domains of SF-12v2, except for SF and RE, and could be predicted by approximately 50%. EQ-VAS was predicted by about 40% using with PF, BP, GH and VT. Both EQ-5D and EQ-VAS were also significantly associated with PCS and MCS.

Table 5

Analysis of EQ-5D, EQ-VAS, Physical and Mental Component Summary scores based on patient characteristics

Table 6

Relationships between eight domains and two component summary scores of SF-12v2 and EQ-5D and EQ-VAS