Disease-related factors, such as the type and severity of disease, the presence of an inhibitor, and hepatitis C virus (HCV) and HIV infection history were obtained from the medical records of the study patients. Unlike other studies, patients’ hemarthropathy status was evaluated by the degree of a patient’s perception of involved major joints including knees, ankles and elbow joints, to assess hemarthropathy from a patient’s point of view. This is because medical records were not available for hemarthropathy evaluation. If the patient felt that the involved joint was very severely affected, a score of 3.0 was allocated. If the patient felt that the joint was less severely affected or not severely affected, a score of 2.0 or 1.0 was allocated, respectively. Subjective joint scores were obtained for each patient after adding together the scores for each affected joint. The analgesic use scores were calculated from the questionnaires. If the patient used analgesics greater than 3 times a week on average, a score of 2.0 was allocated. If the patient used analgesics less than 2 times a week or had no history of analgesic use, the scores of 1.0 and 0.0 were allocated, respectively. Patient disability levels were obtained from physical examinations during the summer camp period by physicians based on the disability level evaluation guidelines in Korea.[18] A lower score meant more severe disability. As a treatment factor, self-injection score was calculated from the questionnaires for each patient. The self-injection score was 3.0 if a patient was able to selfinject coagulation factor concentrates (CFCs) in 80-100% of all moderate bleeding episodes, 2.0 if it was possible in 50-80% of all moderate bleeding episodes, and 1.0 if it was possible in less than 50% of the bleeding episodes. If a patient was not able to self-inject CFCs, a score of 0.0 was allocated. The impact of prophylactic treatment could not be evaluated, because only on–demand treatment had been provided to the patients who were involved in the study, due to national health insurance restrictions on the administration of prophylactic treatment in Korea.

Socio-demographic data, such as marital status, occupational status, and education years, were obtained by questionnaires. The number of work or school day absences per year was also evaluated by questionnaires.

SF36 (standardized questionnaires of the medical outcomes study short form 36 health survey) were used to evaluate QoL of the hemophiliac patients.[19,20] SF36 is composed of 36 questions and eight scales to assess the general health state of each study participant. The eight scales represent physical functioning (PF), role-physical (RP), body pain (BP), general health (GH), vitality (VT), social functioning (SF), role-emotion (RE) and mental health (MH) of each subject, and each scale can have a value ranging from 0 (worst) to 100 (best). The physical component summary (PCS) represents four scales of physical states (physical functioning, role-physical, body pain, general health), and the mental component summary (MCS) represents the other four scales.[20]

Observed disease-related factors, a treatment factor and socio-demographic data were compared among the patient groups that were divided by the severity of hemophilia and the presence of inhibitor. The QoL determined using SF36 was compared among the patient groups, and was compared with the published data for the QoL of osteoarthritic patients and normal healthy control subjects in Korea.[21] Paired t-test was used for comparisons, and a P-value less than 0.05 was defined as significant. In addition, correlation analysis and stepwise multiple regression analysis were conducted between the factors (disease-related factors, the treatment factor and socio-demographic data) and each scale of the SF36 using SPSS version 12.0 (SPSS Inc., Chicago, IL, USA). Because the QoL generally declines with age[22] and hemarthropathy also worsens with age,[23] age-adjusted analysis was conducted.

Initially, 108 patients participated in this study. However, only 55 hemophiliac patients completed all of the questionnaires and these patients’ clinical data was collected by reviewing their medical records. There were 35 severe hemophiliac patients (64%), 5 moderate hemophiliac patients (9%) and 15 inhibitor patients (27%). There was no mild hemophiliac patient. Most (80%) of the inhibitor patients were high responding inhibitor patients (Table 1).

Seventy one percent of patients were already disabled, and 19% of patients were regularly using analgesics. The average subjective joint score was 6.76, which meant that on average, the study patients felt very severe joint abnormalities in at least two major joints. There were no HIV infected patients, but 36.3% of the patients were infected with HCV. Ninety percent of the patients could self-inject CFCs for over 50% of all moderate bleeding episodes, and the average self-injection score was 1.71±0.63. The average age and years of education of the patients were 33.2±12.9 years and 13.06±2.2 years, respectively. No subject under 24 years of age was married or employed. Fifty percent of the patients over 24 years of age were married, and 44% of the patients over 24 years of age were employed. The average loss in work or school was 6.6 days in a year (Table 1).

There was no difference in the average age among the following three groups- severe hemophiliac, moderate hemophiliac, and inhibitor patient groups. Compared with the severe hemophiliac patient group, the subjective joint scores were nearly twice (1.8 times) as high, analgesic scores were 3.2 times higher and work- day loss was 1.6 times higher in inhibitor patients. In contrast, disability scores, self-injection scores and socio-demographic data were not significantly different between the severe hemophiliac and inhibitor groups. The number of work days lost and analgesic use were lower and the percentage of patients who were actively working was higher in the moderate hemophiliac patient group compared with the severe hemophiliac and inhibitor patient groups (Table 1). Statistically the QoL of inhibitor patients were significantly lower in the BP, PCS and MH scales compared with severe hemophilia patients (Fig. 1, ★ P<0.05). The QoL was not different between the severe and moderate hemophiliac patients, except for BP (Fig. 1, ◆ P<0.05).

When the QoL was compared according to differences in the sociodemographic state, there were no significant differences in the QoL between married and unmarried patients and patients with a job and without a job. There were significant differences in some scales of SF 36 above a certain value of age, subjective joint score, disability score, and analgesic score of those patients. In patients older than 24 years of age, the PF, GH, VT, SF, MH, and PCS, MCS scores were significantly lower when compared with those younger than 24 years of age. In patients having the subjective joint score ≥ 4.0, the BP, GH, VT, SF, MH, and PCS scores were significantly lower when compared with those having the subjective joint score<4.0. In patients with the disability score≥ 3.0, only PF score was significantly higher when compared with those having the disability score<3.0. In patients with the analgesic use score≥ 1.0, the PF, BP, GH, VT, SF, MH, and PCS scores were significantly lower when compared with those having the analgesic use score<1.0 (Table 2).

The age-adjusted partial correlation analysis showed that the severity of disease and inhibitor presence was negatively correlated with BP, PCS, SF, and PF. The analgesic use score and the disability level were correlated with several scales of SF36. However, neither the subjective joint score nor the presence of HCV infection showed any significant correlation with any of the SF36 scales (Table 3). The analgesic use score was negatively correlated with PCS, BP, and VT. The disability score was correlated with PF and GH (Table 3). However, when stepwise multiple regression analyses were conducted for the SF36 scales, which correlated with the disease- related factors, severity of disease and inhibitor presence were the significant variables only for BP (β=0.752, P=0.048), the analgesic use score was only a significant variable for PCS (β=-0.695, P=0.002), and the disability level was not a significant variable for any of the correlated SF36 scales (Table 4). In the socio-demographic data, only marital status was correlated with the SF36 scales; it was correlated with PCS, GH, PF, BP, VT, and SF when age-adjusted correlation analysis was conducted (Table 3). However, stepwise multiple regression analysis for these correlated SF36 scales revealed that marital status was a significant variable only for the physical SF36 scale, BP (β=0.638, P=0.02) (Table 4). The self-injection score is the only treatment factor that could be surveyed in the study and significantly correlated with GH, SF, and MH (Table 3). The stepwise multiple regression analysis for these SF36 scales showed that self-injection score was a significant variable for all these SF36 correlated scales [GH (β=0.895, P=0.011), SF (β= 0.489, P= 0.046) and the MH domain (β=0.611, P=0.040)] (Table 4)

The marital status and years of education completed by the Korean patients with hemophilia were not different from those of normal healthy subjects of a similar age group (Table 5). However a significantly lower proportion of hemophiliac patients were in the active work force (44% of hemophiliac patients vs. 100% of healthy individuals). The QoL of hemophiliac patients was much lower than that of normal control subjects in all SF36 scales, including the mental health scales; the average scores for each scale ranged from 44.1% (RP)-74.8% (VT), compared to those of the control subjects (Fig. 2). The hemophiliac patients had similar QoL to osteoarthritic patients in Korea, especially in the physical health scales. However, the average age of the osteoarthritic patients was nearly twice that of the hemophiliac patients, and the SE36 mental health scales (SF, RE, MH and MCS) of the osteoarthritic patients was better than that of the hemophiliac patients (Fig. 2).

The age-adjusted partial correlation analysis showed that the severity of disease, presence of inhibitor, analgesic use and disability were associated with the SF36 physical and mental health scales, in the present study (Table 3). However, the stepwise multiple regression analysis showed that all of these factors were significant variables only for a few physical health domains of SF36 (Table 4). This was very similar to the results observed in other studies; the severity of disease,[7-9] inhibitor presence,[24,25] and arthritis[9,11-13] correlated only with the SF36 physical health scales. Unlike other studies, our study patients’ subjective report of their joint state was shown to not correlate with their QoL. This result may be derived from the limitations of arthropathy evaluation in this study because most of the medical records were not available to appropriately evaluate hemarthropathy. Consequently, we could not objectively determine the state of arthropathy. Instead, analgesic use and disability, which may represent a state of more severe hemarthropathy, were associated with QoL. According to the study conducted by Fischer et al.[13], the impact of hemarthropathy on the QoL in hemophiliac patients may not be significant unless the patient is severely affected. Therefore, the subjectively evaluated joint state may have had an impact on the QoL only after the hemarthropathy was severe enough to necessitate the use of analgesics, or which would have resulted in disability. This speculation was supported by the observation of significantly lower values of several physical and mental health scales of those patients whose subjective joint scores were above 4.0 points when compared to patients with subjective joint scores below 4.0 points (Table 2). In addition to the impact of disease-related factors on the QoL, social status (marital status) and treatment factor (self-injection score) were correlated with the SF36 physical and mental health scales (Table 3). Marriage may have a positive impact on the QoL of hemophiliac patients; a study conducted in Austrian patients showed that marriage may be the cause of nearly normal emotional and mental health scores in Austrian hemophiliac patients.[26] However, there were no differences in emotional and mental health scores between the married and unmarried patients in the present study (Table 2), and marital status was only a significant variable for the physical domain (BP: β= 0.638, P= 0.02) of SF36, based on stepwise multiple regression analysis (Table 4). Therefore, marital status may have little impact on the mental health of these Korean patients. In contrast, the ability of these patients to self-inject was the only factor that had an impact on the mental health of hemophiliac patients. The self-injection score was the only treatment factor that could be recorded in the study. The mental health scale (MH) and two other scales (GH and SF) that represent physical and mental health of patients[20] were correlated with self-injection ability when the age-related partial correlation analysis was conducted (Table 3). Furthermore, the self- injection score was shown to be significantly correlated with the SF36 scales when stepwise multiple regression analysis was conducted (Table 4). Few studies have been conducted about the impact of self-injection on the QoL of hemophiliac patients. The ability to self inject may have a significant impact on the physical, social and mental health status of hemophiliac patients because these patients can rapidly control their bleeding, and feel that they can independently control their disease and actively participate in social activities. As a result, the ability to self- inject may have an impact on both the physical and mental health of hemophiliac patients.

Mental health has been shown to be poorer in Korean hemophiliac patients compared with age-matched normal healthy subjects and osteoarthritic patients in Korea (Fig. 2). This result was different from that of other studies, most of which were conducted in developed countries that have a long history of well-organized hemophilia care systems.[6-8] Results of these studies showed that the mental health scale scores were similar to the normative means.[6-8] Social status was not different between hemophiliac patients and normal healthy subjects of similar ages in the present study except for employment rate, which was higher in normal healthy subjects (44% of hemophiliac patients vs. 100% of healthy individuals) (Table 5). Markova et al.[27] showed that unemployed hemophiliac patients appeared to be less adjusted to life and suffered most from psychosomatic symptoms. However, the lower employment rate could not alone explain the lower mental health among Korean hemophiliac patients. The mental health scales score for hemophiliac patients was similar to that of normal control subjects, although the employment rate in Austrian patients[26] was also significantly lower than normal control patients (64% of hemophiliac patients vs. 90% of healthy individuals). The proportion of patients with different hemophilia severities in this study were different from other studies.[7,8] There was no mild hemophiliac patient, and only 9% of the patients had moderate hemophilia. 64% of the patients had severe hemophilia, and 27% of the patients had inhibitors present, most of whom were high responding inhibitor patients (Table 1). The difference in the proportion of hemophilia severities may explain the different scores of mental health scales and the lower scores in all SF36 scales compared with other studies.[7-11,24,25] However, mental health was not different between patients with various disease severities in these studies. Therefore, the poorer mental health in Korean hemophiliac patients could not be explained by the different proportion of disease severities in hemophiliac patients. The social, cultural, physical and educational status may all affect a person’s mental health.[28] However, according to the result of a study conducted in Chinese hemophiliac patients,[29] mental well being was not correlated with hemarthropathy, salary, or cost of replacement therapy. Similar results were also observed in Korean hemophiliac patients. Neither disease-related factors nor social status were significantly associated with the mental health of these patients when multiple regression analysis was conducted (Table 4). Only the self-injection score, which was the only treatment factor that could be surveyed in this study, had a significant impact on the mental health of Korean hemophiliac patients (Table 4). Few researchers have studied the impact of treatment factors on the mental health of hemophiliac patients. A comparative study showed no difference in the QoL between patients treated with prophylaxis and patients treated with on-demand treatment.[6] However, there are many treatment factors that may affect the QoL of hemophiliac patients, including the ability to self- inject, as well as the presence or absence of well-organized hemophilia treatment centers, resource constraints and the cost of hemophiliac therapy, which vary in different countries. Therefore the differences in treatment factors across countries may contribute to poorer mental health in hemophiliac patients in Korea and China, although the etiology remains to be elucidated.

In conclusion, disease-related factors and social status only had an impact on the physical health of Korean hemophiliac patients. However, the ability to self-inject had an impact on both the physical and mental health of Korean hemophiliac patients. In contrast to the results found by studies carried out in developed countries, mental health was significantly poorer in Korean hemophiliac patients. Differences in the treatment factors across countries may have contributed to the poorer mental health of Korean hemophiliac patients.