Health related quality of life in sickle cell patients: The PiSCES project

Functional status and health-related quality of life (HRQOL) may be impaired in sickle cell disease (SCD) due to morbid events, such as stroke, or other organ system failures. The Cooperative Study of Sickle Cell Disease (CSSD) found that morbid events such as strokes that impaired function often preceded death in childhood [1‐3] Until recent decades, SCD was associated with chronic childhood pain, organ failure and death in very early adulthood. Treatment advances have now transformed SCD into a chronic disease suffered by children and adults. Frequently, patients surviving until adulthood experience significant organ system damage that may include stroke, pulmonary failure and pulmonary hypertension, renal failure, congestive heart failure, leg ulcers, and osteonecrosis of the femoral or humeral heads [2].

Children and adolescents with SCD report poor HRQOL in qualitative studies using focus groups [4], and fare worse in their HRQOL compared to controls on health surveys [5] or on assessments of general physical, motor and independent daily functioning [6, 7]. Despite the considerable evidence in children for reduced HRQOL in SCD, few studies have evaluated the impact of this disease on health related quality of life in adults [8‐11].

The impact of this disease on HRQOL for adults may be even greater than for children. Quality of life is deteriorated by episodic, debilitating pain associated with substantial analgesic use, frequent hospitalization for pain episodes, and ultimately organ failure. Although SCD related pain can often be managed by analgesics and opioids, adults with SCD may be under-treated because clinicians suspect drug dependence in this population. This provider bias may lead to reluctance by patients to seek medical attention [12, 13]. Further, HRQOL may be over-estimated by providers that do not regularly care for patients with SCD due to lack of understanding of the severity of the painful crises and the potential impact on function. Therefore, assessing the quality of life among a US adult SCD population and providing comparisons with HRQOL reported among comparable adults with other chronic diseases may be an important method for providing health care practitioners who care for these patients a more objective perspective on the impact and severity of this disease. In addition, quality of life in SCD is important to describe because it amplifies the ability to identify patients for whom potentially dangerous but potent interventions such as hydroxyurea or bone marrow transplantation is justified at an early stage.

Table 1 describes the PiSCES cohort. The mean age was 33, and ranged from 16 to 64. There were more women than men in the study (60.4% vs 39.6), 48.6% attended college. Only 21.8% of subjects were currently married.

Table 2 has means and standard deviations for the 8 SF-36 subscales for the PiSCES cohort, separately for men and women, along with the age-adjusted national norms by gender. When the gender stratified PiSCES cohort was compared with national norms, values were significantly lower for all subscales (P < 0.0001) with one exception – the mental health scale was not significantly different from the national norm (men: p = 0.670, women: p = 0.102).

Table 3 has means and standard deviations of the 8 SF-36 subscales for the PiSCES cohort along with cohorts of patients with three other chronic diseases – asthma, cystic fibrosis and hemodialysis. There were significant differences (P < 0.0001) amongst cohorts for all subscales except mental health (p = 0.0582), which was marginal. Patients with SCD (PiSCES cohort) reported significantly worse HRQOL on all subscales (p < 0.05) except mental health as compared to adolescents and adults with cystic fibrosis. They had similar reported quality of life as asthma patients regarding physical function and role function (both physical and emotional), and mental health, but scored lower for the bodily pain, vitality, social function and general health subscales. When compared to patients on hemodialysis, SCD patients reported similar low scores for physical and emotional role function and social function. They also did not differ on the mental health subscale (p > 0.15). SCD patients had lower scores for the pain, vitality and general health subscales (P < 0.01), but reported a higher score for the physical function subscale compared with the hemodialysis cohort.

Multiple regressions were performed to look at the relationship of SCD specific variables (genotype, mean pain, percent of diary days subjects reported crisis and percent of diary days subjects reported utilization) and SF-36 subscales. Socio-demographic variables (age, gender, number of years of education) were included in the models as covariates. Results are in Table 4. For each subscale, mean SCD pain was highly predictive (p < 0.0001 for all subscales except p = 0.0396 for mental health). The more SCD pain a subject experienced, the worse the reported quality of life. One unit increase in pain (on a 0–9 scale) was associated with an approximate decrease of 1.4 (mental health) to 6 (both role functions) units on an SF-36 subscale. Percentage of days with crisis was an independent predictor of bodily pain (p = 0.0109), with an approximate 6 point decrease in bodily pain score for each increase in crisis category. Genotype was also an independent predictor of vitality (p = 0.0161), with SS/Sß^o Thalasemia being associated with better vitality. No other variables were independent predictors of SF-36 subscales.

In general, SCD patients experience a poor health related quality of life. Except for mental health, the SF-36 subscale values were considerably lower than norms of the general US population. They reported a HRQOL that was equal to or poorer than patients with other significant chronic conditions in many domains. Similar to patients with SCD, until somewhat recently, patients with cystic fibrosis rarely lived until adulthood, marking this as a disease with significant sequelae and high mortality. It is interesting, then, to see that quality of life of adult survivors of this chronic disease, even though impaired, was comparable to national norms, and was generally far superior than that reported by adults with SCD. Of the three chronic conditions selected for comparison, the PiSCES cohort had HRQOL patterns most similar to that of patients undergoing chronic hemodialysis.

That SCD patients did not report poorer mental health and well being than the general US population is consistent with findings for many medical conditions. In 1978 Brickman, et al [26] presented a famous result showing that lottery winners are not much happier than paraplegics. Since that time many studies have confirmed similar results, that while people who are sick may report their health as being worse than the general population, they appear to have similar well-being. Not only was it true of the asthma, dialysis and cystic fibrosis cohorts presented here, but similar results have been shown for people with other chronic diseases both in the US and other countries [21, 22, 27].

It has been suggested that the fact that many people with chronic diseases report good psychological well being could be a result of increased social support, lack of other stressors, or a "response shift" associated with the managing their chronic disease [27]. The "response shift" could be a result of a scale recalibration, a change in the patient's values, or a reconceptualization of their mental health and well-being [28, 29] in order to accommodate their illness. Riis et al [30] dispute the idea of scale recalibration, proposing instead that people have adapted to their illness or situation.

Heady and Wearing [31] propose that there is a baseline level of mood or well-being that people have to which they return after events cause them to move from that baseline. This is supported by a twin study indicating that most variation in well-being is due to variations in genetics, not life circumstances [32]. That would suggest that, while perhaps people with SCD may temporarily report poorer well-being associated with high levels of pain or other disease sequalae, most often they would report a baseline well-being similar to that of others.

Despite the similar level of well-being in SCD patients compared with both norms and patients with other chronic diseases, patients with SCD experience significant decrements in other important aspects of HRQOL. This is supported in a study of adult SCD patients in the UK, where Anie et al, found their population also had much lower HRQOL scores than general UK population norms. Further, the patients in this study had reported HRQOL similar to patients with arthritis due to hereditary haemochromatisis, another chronic disease [9]. Patients in the PiSCES cohort reported somewhat lower general health and higher mental health scores than Anie et al found. This may result in part from differences in the two cohorts, including their relative access to health care in the two settings.

Surprisingly, HRQOL was not associated with genotype except for the vitality subscale. However there was a strong association with reduced HRQOL and pain levels, and, for a few subscales, there was a trend with increasing levels of utilization. The relationship between genotype severity and HRQOL may have been mediated by these variables, particularly pain. Anie et al [9] also found a relationship between pain and some subscales of the SF-36 (physical and social functions, mental and general health) but found no significant association with utilization measures. Whether the more severe manifestations of the disease cause the poorer quality of life, or patients who report poorer quality of life suffer more and use more health care cannot be determined from this study.

It is unclear why the direction of the statistically significant association between vitality for SS/Sß^o Thalasemia vs SC/Sß⁺ Thalasemia is opposite of what would be expected, with the more severe genotype being associated with better vitality. Even the nonsignificant relationships between genotype and the other HRQOL subscales in these regressions were in the same direction, except for physical functioning. Similarly, increased utilization tended to be associated, albeit nonstatistically with better HRQOL scores for most subscales These counter-intuitive findings should be explored by other SCD researchers.

There are several limitations of this study. First, this study enrolled patients from only one state, so may not be representative of the entire US SCD population. When comparing to populations with chronic disease, the gender and age distributions were not ideally matched. In particular, the PiSCES cohort, at a mean age of 33, is significantly younger than the dialysis comparison group However, since SF-36 subscale scores tend to decrease with age, the fact that these younger patients with SCD had worse HRQOL scores on some subscales than an older hemodialysis cohort is even more alarming

Practitioners caring for adult SCD patients should regard their quality of life as severely compromised, with scores that are most similar to hemodialysis patients in our comparison of other chronic diseases. Although reducing mortality is of paramount importance among SCD patients, future interventions should consider improving health related quality of life as a clinical endpoint.