SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Objective

To evaluate both direct and indirect costs of systemic lupus erythematosus (SLE) patients with and without flares from a societal perspective, and to investigate the impact of the severity and clinical manifestations of flares on direct/indirect costs.

Methods

A retrospective cost-of-illness study was performed on 306 SLE patients. Participants completed questionnaires on sociodemographics, employment status, and out-of-pocket expenses. Health resources consumption was recorded by chart review and patient self-reported questionnaire. The total number of flares and involved organs during the preceding 12 months were recorded. Multiple linear regression was performed to determine the cost predictors.

Results

Patients with flares were younger, had shorter disease duration, and had higher disease activity at the time of the assessment. The overall incidence of lupus flares was 0.24 episodes per patient-year. Patients with flares used more health care resources and incurred significantly higher annual direct and indirect costs. The mean total costs per patient-year were 2-fold higher for patients with flares ($22,580 versus $10,870 [2006 US dollars]; P < 0.0005). Multiple regression analysis showed that the number of flares was an independent explanatory variable associated with increased direct costs. Patients with multiorgan flares or renal/neuropsychiatric flares incurred higher direct costs compared with those with single-organ flares or with other organ flares.

Conclusion

Patients with flares incur higher direct and indirect costs compared with those without flares. Major organ flares incur higher disease costs than other organ flares. Treatments that effectively control disease activity and prevent flares, especially major organ flares, may reduce the high costs associated with flare in SLE.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Systemic lupus erythematosus (SLE) is a chronic, multisystem, autoimmune disease characterized by periods of fluctuating disease activity. The British Isles Lupus Activity Group (BILAG) index (1) and the Safety of Estrogens in Lupus Erythematosus: National Assessment (SELENA) version of the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) (2) are 2 disease activity indices primarily used in the clinical studies of SLE. The assessment of lupus activity encompasses the concept of a flare, which is an increase in disease activity over a defined period (3). However, there is no general consensus on the definition of flare, although various tools have been used (4–7). Using an increase of 1.0 cm on a 3-cm visual analog scale of the physician's global assessment (PGA) of disease activity as a gold standard of flare, the corresponding cutoff is 3 points or more on the SELENA–SLEDAI and 4 points or more on the BILAG (8). Since indices alone may not capture overall changes in activity, SELENA trial investigators developed the SELENA flare tool, which incorporates 2 indices of disease activity (PGA and SELENA–SLEDAI), clinical manifestations, and treatment to define both mild/moderate and severe flares (2).

Flare is an important outcome variable and has been shown to be a major cause of admission (9). Disease activity and toxicity of the consequent treatments result in irreversible damage that is associated with an increased risk of morbidity and mortality (10). Stoll et al concluded that death and the long-term accumulation of damage were strongly predicted by a high total disease activity over time, and especially associated with the number of BILAG A (most active disease) flares (11). Contrary to the so-called minor organ flares, i.e., constitutional, musculoskeletal, and mucocutaneous (4), major organ flares such as renal or neuropsychiatric (NP) flares have been shown to be associated with poor prognosis. Renal flares were significantly associated with the risk of doubling plasma creatinine level and death or dialysis (12). Ward et al concluded that the occurrence of seizures increased the risk of death in patients with SLE (13). Results from Hanly et al showed that NP disease was related to more frequent use of corticosteroids and immunosuppressants (14).

Previous studies on the economic impact of SLE focus on the relationship between disease activity/damage and costs. Higher disease activity/damage has been shown to be associated with both higher direct and indirect costs (15–18). However, to our knowledge, no study has focused on the relationship between costs and flares. Whether the severity or specific clinical manifestations of flares would influence disease costs has not been studied.

In the current study, we evaluated both direct and indirect costs of SLE patients with and without flares from a societal perspective. We also investigated the impact of the severity and clinical manifestations of flares on direct/indirect costs. In view of the evidence that major organ flares are significantly related to poor prognosis, we selected 2 major organ flares, renal and NP flares, to find out whether major organ flares were more costly than other flares.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Patients and procedures.

A convenience sample of 306 Chinese patients with a diagnosis of SLE according to the 1997 American College of Rheumatology (ACR) revised criteria (19), who had been followed at the Rheumatology Clinic of the Prince of Wales Hospital in Hong Kong, were recruited between January 2006 and August 2007. All of the participants were within working age (≥18 years; <65 years for men and <60 years for women) and were followed at the Prince of Wales Hospital at regular intervals (every 3 to 4 months) according to a standardized assessment protocol, including 1) disease activity assessment according to the SELENA–SLEDAI at each visit (20) and 2) yearly disease damage assessment according to the Systemic Lupus International Collaborating Clinics/ACR Damage Index (SDI) (21). Patients who were not capable of responding to a questionnaire (e.g., presence of dementia) were excluded. The Ethics Committee of the Chinese University of Hong Kong approved this study, and all of the patients provided written informed consent.

A questionnaire including sociodemographics, employment outcomes, and patients' out-of-pocket expenses was administered by a trained interviewer. The same questionnaire had been used in a cost-of-illness study of patients with ankylosing spondylitis (22). Clinical and laboratory assessments were also performed in all of the subjects by their treating rheumatologists, including the SELENA–SLEDAI and the SDI. The SELENA–SLEDAI is a valid and reliable disease activity measure of SLE (20) that contains 24 descriptors in 9 organ systems, including clinical and laboratory measures. The total SELENA–SLEDAI score falls between 0 (no activity) and 105 (maximum activity). The SDI, a validated physician-rated index that consists of 41 items in 12 organ systems/domains, was used to measure accumulated damage (21). Damage was defined as any irreversible change occurring since the onset of SLE and presenting for at least 6 months. The total SDI scores range from 0 (no damage) to 49 (maximum damage).

Definitions of flare.

Patients' medical records were then reviewed by an investigator (TYZ) to derive the total number and manifestations of flares during the preceding 12 months. A revised SELENA flare tool that excluded the component of PGA was used to define flares (2). Mild/moderate flares were defined as 1 or more of the following: 1) change in SELENA–SLEDAI score of >3 points but ≤12 points; 2) new/worse discoid lesions, photosensitivity, profundus, cutaneous vasculitis, bullous lupus, nasopharyngeal ulcers, pleuritis, pericarditis, arthritis, and/or fever (SLE); 3) increase in prednisone not to exceed 0.5 mg/kg/day; or 4) added nonsteroidal antiinflammatory drugs (NSAIDs) or hydroxychloroquine for SLE. Severe flares were defined as 1 or more of the following: 1) change in SELENA–SLEDAI score of >12 points; 2) new/worse NPSLE, vasculitis, nephritis, myositis, platelet count <60,000/mm3, or anemia (hemoglobin level <7 mg/dl), which required either a doubling of or increase in prednisone dosage to >0.5 mg/kg/day; 3) increase in prednisone to >0.5 mg/kg/day; 4) new immunosuppressants for SLE activity; or 5) hospitalization for SLE.

The definitions of the individual organ flares are listed below. Renal flare was defined as 1 of the following (23, 24): 1) a reproducible (2 samples at least 1 week apart) increase in 24-hour urine protein levels to >1 gm if the baseline value was <0.2 gm, to >2 gm if the baseline value was 0.2–1 gm, or to more than twice the baseline value if the baseline value was >1 gm; 2) a reproducible increase in serum creatinine level of >20% or at least 25 μmoles/liter, whichever was greater, accompanied by proteinuria (>1 gm/24 hours), hematuria (≥4 red blood cells [RBCs]/high-powered field [hpf]), and/or RBC casts; or 3) new, reproducible hematuria (≥10 RBCs/hpf) or an increase in hematuria by 2 grades compared with baseline, associated with >25% dysmorphic RBCs, exclusive of menses, accompanied by either a 0.8-gm increase in 24-hour urinary protein levels or new RBC casts.

NP flare was defined according to the case definition system for central nervous system lupus syndromes by the 1999 ACR nomenclature (25). This includes a detailed glossary and diagnostic guidelines for 19 NP syndromes, namely aseptic meningitis, cerebrovascular disease, demyelinating syndrome, headache, movement disorder, myelopathy, seizure disorder, acute confusional state, anxiety disorder, cognitive dysfunction, mood disorder, psychosis, Guillain-Barré syndrome, autonomic neuropathy, mononeuropathy (single/multiplex), myasthenia gravis, cranial neuropathy, plexopathy, and polyneuropathy.

Other clinical features of flares were grouped into the following organs/systems: musculoskeletal, mucocutaneous, hematologic, vasculitic, and serositis. Each organ/system flare was defined according to the definitions of the descriptors of the SELENA–SLEDAI instrument (2, 20) (for details, see Supplemental Appendix A, available in the online version of this article at http://www3.interscience.wiley.com/journal/77005015/home). Flares with only serologic manifestations (increased anti–double-stranded DNA [anti-dsDNA] titer and depressed complement levels) without medical intervention were not included in the analyses.

Single-organ flares referred to flares involving only 1 organ/system, whereas multiorgan flares involved more than 1 organ/system (excluding immunologic manifestations).

Cost assessment.

Hong Kong's health care system is largely administered by the government, which provides all residents with comprehensive health care from primary to tertiary care, including medications, investigations, ambulatory care, hospitalization, and operations. The public is charged nominal fees for medical treatment provided by the government. Government hospitals and clinics deliver most of the medical services, especially specialty clinics and inpatient care, with a market share of more than 90% (26, 27). There are also private hospitals in Hong Kong that are run on a profit basis. They are relatively small in number, size, and custom, and used mainly by expatriates and wealthy Chinese. Fees and charges of private hospitals vary considerably (27). In Hong Kong, patients with chronic diseases mainly rely on government hospitals, whereas use of private hospital services represents a relatively small percentage (28). We recorded both government and private medical services by different methods. Use of private hospital facilities was reported by the patients. Use of government hospital services was derived by chart review. We used average per diem costs (both hospital and ambulatory services) estimated by the government authority as a measure of costs of both government and private medical services. The unit costs of some major services have been described elsewhere (22).

We assessed both direct and indirect costs from a societal perspective. Details relating to direct costs were collected for the previous 12 months, consisting of 1) all visits to health care providers, 2) all diagnostic examinations, 3) medications taken, 4) emergency room visits, 5) costs of inpatient care (including rehabilitation hospitalization), 6) costs of private hospital/clinic facilities (including costs of visits, medications, investigations, and hospitalizations), and 7) patients' out-of-pocket expenses for health products, nontraditional therapies (hydrotherapy, acupuncture, and massage), aid devices, transportation fee to the health care providers, private household helper, and adaptation to houses.

Indirect costs represented the productivity loss due to SLE, which included annual sick leave due to SLE, unemployment due to SLE, and days off from household work or daily activities due to SLE. In the mentioned questionnaire, participants were asked to indicate 1) sick leave taken in the preceding 12 months (for those who were still employed), 2) whether they were unemployed due to SLE and the duration of unemployment (for those who were unemployed), and 3) the number of days off from household work or daily activities due to SLE. The human capital approach, which uses wages as a proxy measure of the output of work time to value the individual's lost work hours, was used to calculate productivity loss (29). In our study, wages were derived from Wage and Payroll Statistics, Census and Statistic Department of Hong Kong.

Statistical analyses.

Statistical analyses were performed using the Statistical Package for the Social Sciences for Windows, version 13.0 (SPSS, Chicago, IL). Results are expressed as the mean ± SD for normally distributed data. For non–normally distributed data, the median and interquartile range (IQR) are expressed. A 2-sample t-test, chi-square test, or Mann-Whitney U test was used to compare sociodemographics, clinical features, health care resource use, and disease costs between patients with and without lupus flares. P values less than 0.05 were considered significant. A Kruskal-Wallis test was used to test for differences in costs between patients grouped by severity/organ involvement/manifestations of flare. When a Kruskal-Wallis test revealed significant results, a Mann-Whitney U test by Bonferroni adjustment was used for multiple comparisons (for triple comparisons, P values less than 0.01 were considered significant). Stepwise multiple linear regression analysis was used to determine the cost predictors. Log10 transformation of costs was performed to fit the normative assumptions. The possible cost predictors included the patient's age, education level, disease duration since diagnosis, SELENA– SLEDAI, SDI, and the number of flares during the past 12 months. A sensitivity analysis was performed to indicate whether the test was sensitive to outliers (a case was an outlier if it was 3 SDs away from the mean).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Sociodemographics and clinical profiles.

Table 1 shows the sociodemographics and clinical characteristics (ever) at the time of the assessment of the whole cohort, as well as the 2 groups subdivided according to whether they experienced a flare in the preceding year. Compared with those without flares, patients with flares were younger, had a shorter disease duration, and had higher disease activity at the time of the assessment. Regarding the clinical features, patients with flares had a higher prevalence of having had discoid lesions and being anti-dsDNA positive. No significant differences in the prevalence of major organ manifestations and the SDI score were observed between the 2 groups.

Table 1. Baseline sociodemographics and clinical characteristics (ever) of patients with and without flares in the preceding year*
 Without flares (n = 244)With flares (n = 62)PEntire group (n = 306)
  • *

    Values are the number (percentage) unless otherwise indicated. SELENA–SLEDAI = Safety of Estrogens in Lupus Erythematosus: National Assessment version of the Systemic Lupus Erythematosus Disease Activity Index; SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; anti-dsDNA = anti–double-stranded DNA; ANA = antinuclear antibody.

Age, mean ± SD years42.5 ± 11.436 ± 10< 0.000541 ± 11
Women236 (97)58 (94)0.271294 (96)
Education level, mean ± SD years10.1 ± 4.411.3 ± 3.50.11510 ± 4
Disease duration, mean ± SD years10.2 ± 7.07.4 ± 5.80.0029.6 ± 6.9
SELENA–SLEDAI score, mean ± SD2.15 ± 2.643.63 ± 3.20< 0.00052.5 ± 2.8
SDI score, mean ± SD0.73 ± 1.060.63 ± 1.070.2790.71 ± 1.07
Organ manifestations    
 Malar rash107 (44)26 (42)0.786133 (44)
 Discoid lesion27 (11)13 (21)0.03940 (13)
 Photosensitivity77 (32)20 (32)0.91697 (32)
 Oral ulcer73 (30)21 (34)0.54794 (31)
 Arthritis192 (79)43 (69)0.12235 (77)
 Serositis68 (28)17 (27)0.94485 (28)
 Renal disease139 (57)43 (69)0.076182 (60)
 Neuropsychiatric disease62 (25)21 (34)0.18183 (27)
Hematologic208 (85)57 (92)0.167265 (87)
 Leukopenia124 (51)36 (58)0.308160 (52)
 Lymphopenia161 (66)38 (61)0.489199 (65)
 Thrombocytopenia70 (29)21 (34)0.42591 (30)
 Hemolytic anemia19 (8)6 (10)0.62725 (8)
Immunologic230 (94)60 (97)0.428290 (95)
 Anti-dsDNA positive179 (73)53 (86)0.047232 (76)
 Anti-Sm positive47 (19)19 (31)0.05266 (22)
 Anti-Ro positive134 (55)36 (58)0.656170 (56)
 Anti-La positive50 (21)7 (11)0.09757 (19)
ANA positive241 (99)60 (97)0.801301 (98)

Lupus flare profiles.

During the preceding year, 74 episodes of flare were recorded in 62 (20.3%) of 306 patients. The overall flare rate was 0.24 episodes per patient-year. Fifty (80.6%) of 62 patients had 1 flare and 12 (19.4%) of 62 had 2 flares. Renal flare was the most common (0.09 episodes/patient-year), followed by mucocutaneous flare (0.04 episodes/patient-year), musculoskeletal flare (0.04 episodes/patient-year), hematologic flare (0.04 episodes/patient-year), NP flare (0.03 episodes/patient-year), vasculitic flare (0.03 episodes/patient-year), and serositis flare (0.01 episodes/patient-year). Seven patients had 8 NP flares during the preceding year. Four of 8 were cardiovascular accidents (3 were strokes and 1 was a transient ischemic attack), 2 were seizure disorders, 1 was a migraine, and 1 was a myelopathy.

For those with 1 flare, 18 (36%) of 50 patients had a mild/moderate flare and 32 (64%) of 50 had a severe flare. For those with 2 flares, 1 (8%) of 12 had 2 mild/moderate flares, 6 (50%) of 12 had 1 mild/moderate flare and 1 severe flare, and 5 (42%) of 12 had 2 severe flares. The majority of these patients had a single-organ flare (54 [87%] of 62). Among patients with single-organ flare, 23 (42.6%) of 54 patients had a renal flare, 4 (7.2%) of 54 patients had an NP flare, 10 (18.5%) of 54 patients had a mucocutaneous flare, 8 (14.8%) of 54 patients had a musculoskeletal flare, and 7 (13.0%) of 54 patients had a hematologic flare. Eight (12.9%) of 62 patients had a multiorgan flare involving 2–5 organ systems (median 2). The commonly involved organ systems included the kidney (5 [62.5%] of 8 patients), brain (4 [50%] of 8 patients), hematologic system (4 [50%] of 8 patients), vasculitis (3 [37.5%] of 8 patients), and musculoskeletal system (2 [25%] of 8 patients).

Flare, health care resources utilization, and costs.

Table 2 shows the health care resources use of patients with and without flares. More visits to rheumatologists were observed in patients with flares. Seventy-six percent of patients with flares had urine tests compared with 37% of those without flares (P < 0.0001). The proportion of patients having imaging tests was also higher in those with flares (28% versus 50%; P = 0.001). A higher proportion and longer duration of inpatient care were seen in patients with flares. For those with flares, the major reason for hospitalization was flare (58%), followed by infection (14%). For those without flares, infection was the major reason for hospitalization (30%).

Table 2. Health care resources use with regard to patients with and without flares during the preceding year*
 Without flares (n = 244)With flares (n = 62)P
  • *

    Values are the mean ± SD (median) unless otherwise indicated.

  • A standard clinic consists of a general outpatient department and a family health center, with or without a maternity ward.

  • Including radiographs, ultrasounds, computed tomography scans, and magnetic resonance imaging.

No. of visits to health care providers6.9 ± 4.8 (5)8.8 ± 4 (9)< 0.0001
 Rheumatologist4.23 ± 2.19 (4)6.37 ± 3.07 (5.5)< 0.0001
 Nephrologist0.07 ± 0.57 (0)0 (0)0.257
 Dermatologist0.23 ± 0.93 (0)0.23 ± 0.95 (0)0.888
 Ophthalmologist0.61 ± 1.15 (0)0.52 ± 0.74 (0)0.906
 Government general clinic0.46 ± 1.17 (0)0.85 ± 2.02 (0)0.053
 Allied health0.24 ± 1.93 (0)0 (0)0.149
 Psychologist0.03 ± 0.28 (0)0.06 ± 0.4 (0)0.271
 Others1 ± 2.22 (0)0.74 ± 1.88 (0)0.210
No. of diagnostic examinations   
 Blood test28 ± 45 (24)40 ± 21 (35)< 0.0001
 Urine test4 ± 8 (0)9 ± 12 (3.5)< 0.0001
 Imaging tests0.27 ± 0.74 (0)1.31 ± 2.69 (0.5)< 0.0001
Visit to the emergency room, %2152< 0.0001
No. of visits to the emergency room0.29 ± 0.7 (0)1.03 ± 1.33 (1)< 0.0001
Inpatient care, %1669< 0.0001
Duration of inpatient care, days3.1 ± 16 (0)15.6 ± 37.3 (4.5)< 0.0001

All of the patients with flares required medication treatment in the preceding 12 months, compared with 96% of those without flares (P = 0.105). Use of NSAIDs, corticosteroids, immunosuppressants, angiotensin-converting enzyme inhibitors/angiotensin II receptor blockers, antibiotics, and prophylaxis for steroid-induced osteoporosis was more common in patients with flares (Table 3). The use of health products, nontraditional therapies, aids, and private hospital/clinic facilities did not differ between patients with and without flares.

Table 3. Medications taken (ever) by patients with and without flares in the preceding year*
 Without flares (n = 244)With flares (n = 62)P
  • *

    Values are the number (percentage). NSAIDs = nonsteroidal antiinflammatory drugs; ACE = angiotensin-converting enzyme; ARBs = angiotensin II receptor blockers.

  • Including diuretics, anti-arrhythmic drugs, β-adrenoceptor–blocking drugs, hypertension and heart failure drugs (excluding ACE inhibitors/ARBs), nitrates, calcium-channel blockers, anticoagulants and protamine, antiplatelet drugs, lipid-regulating drugs, fibrinolytic drugs, antifibrinolytic drugs, and hemostatics.

  • Including antacids and simethicone, antispasmodic drugs, ulcer-healing drugs, adsorbents and bulk-forming drugs, antimotility drugs, laxatives, local preparations for anal and rectal disorders, and drugs affecting intestinal secretions.

  • §

    Including hypnotics and anxiolytics, antipsychotic drugs, antipsychotic depot injections, antimanic drugs, antidepressant drugs, drugs for nausea and vertigo, analgesics, antiepileptics, drugs for dementia, and drugs used in parkinsonism and related disorders.

NSAIDs91 (37)33 (53)0.023
Anti-malaria drugs134 (55)35 (57)0.828
Corticosteroids163 (67)60 (97)< 0.0001
Immunosuppressants62 (25)34 (55)0.011
ACE inhibitors/ARBs70 (29)30 (48)0.003
Anti-osteoporosis117 (48)46 (74)< 0.0001
Antibiotics46 (19)19 (31)0.043
Gastrointestinal drugs95 (39)36 (58)0.007
Cardiovascular system drugs77 (32)19 (31)0.890
Neuropsychiatric drugs§38 (16)13 (21)0.309

The employment rate was not significantly different between these 2 groups (46% for patients without flares and 48% for those with flares; P = 0.948). For those who were employed, a higher proportion (93% versus 66%; P = 0.003) and longer duration (median [IQR] 4 [10] days versus 15 [23] days; P < 0.0005) of annual sick leave were observed in patients with flares. There was no difference in the duration of unemployment between the 2 groups for those who were unemployed because of SLE (median duration 12 months for both; P = 0.202). The number of days off from household work or daily activities did not differ between the 2 groups (median [IQR] duration 0 [0] days versus 0 [30] days; P = 0.299).

Patients with flares incurred approximately twice the average annual total costs of those without flares ($22,580 versus $10,870 [2006 US dollars] per patient; P < 0.0005) (Table 4). Annual direct costs were nearly 3-fold higher for patients with flares compared with those without flares (P < 0.0005). Patients with flares incurred significantly higher costs in all of the components of direct costs. For both groups, the costs of inpatient care represented the largest component, accounting for 40% (patients without flares) and 70% (patients with flares) of total direct costs. Annual indirect costs were also significantly higher in those with flares (P = 0.017). For those who were employed, higher indirect costs due to sick leave were also observed in patients with flares. Indirect costs due to SLE-related unemployment and days off from household work or daily activities did not differ between the 2 groups.

Table 4. Annual costs for patients with and without flares (in 2006 US dollars)*
 Without flares (n = 244)With flares (n = 62)
  • *

    1 US dollar = 5.527 Hong Kong dollars. The purchasing power parities conversion factor of 2006 was used and the conversion factor of US dollars is 1:1. IQR = interquartile range.

  • P < 0.005.

  • P < 0.05.

Visits to health care  provider  
 Mean ± SD805 ± 5601,017 ± 440
 Median (IQR)633 (504)1,013 (566)
Diagnostic examinations  
 Mean ± SD1,180 ± 6861,780 ± 1,035
 Median (IQR)989 (636)1,564 (1,232)
Medications  
 Mean ± SD317 ± 515381 ± 404
 Median (IQR)164 (245)265 (322)
Emergency visits  
 Mean ± SD31 ± 79108 ± 140
 Median (IQR)0 (0)103 (206)
Inpatient care  
 Mean ± SD2,425 ± 12,58111,737 ± 25,615
 Median (IQR)0 (0)3,469 (12,759)
Private hospital/clinic  services  
 Mean ± SD48 ± 169115 ± 516
 Median (IQR)0 (0)0 (0)
Patient out-of-pocket  expenses  
 Mean ± SD1,161 ± 2,3291,685 ± 2,649
 Median (IQR)140 (1,210)317 (2,167)
Indirect costs due to  sick leave  
 Mean ± SD1,509 ± 7,3635,014 ± 17,078
 Median (IQR)0 (1,360)0 (5,042)
Indirect costs due to  SLE-related  unemployment  
 Mean ± SD24,201 ± 49,09124,225 ± 49,126
 Median (IQR)0 (0)0 (8,497)
Indirect costs due to  days off from  household work  or daily activities  
 Mean ± SD1,398 ± 9,2732,577 ± 13,002
 Median (IQR)0 (0)0 (0)
Total direct costs  
 Mean ± SD6,034 ± 12,89916,873 ± 25,510
 Median (IQR)2,872 (4,106)9,441 (12,364)
Total indirect costs  
 Mean ± SD4,905 ± 8,8725,756 ± 8,999
 Median (IQR)322 (7,040)1,013 (10,061)
Total costs  
 Mean ± SD10,870 ± 16,09422,580 ± 29,943
 Median (IQR)4,539 (12,689)14,276 (19,423)

In univariate analysis, variables significantly associated with direct costs included age, disease duration, SELENA–SLEDAI score, SDI score, and the total number of flares (Table 5). In multivariate analysis, disease duration, SDI score, and the total number of flares were independent explanatory variables associated with increased direct costs. The total number of flares was the only variable significantly associated with increased indirect costs in univariate analysis (Table 5). However, none of these variables were independent predictors of indirect costs in the multivariate analysis. A sensitivity analysis was performed and the tests were not sensitive to the outliers.

Table 5. Multiple linear regression model of annual direct and indirect costs*
 Univariate analysisMultivariate regression analysis
CoefficientPCoefficientPR2
  • *

    Due to the skewness of direct and indirect costs data, a log10 was performed prior to the regression analysis. SELENA–SLEDAI = Safety of Estrogens in Lupus Erythematosus: National Assessment version of the Systemic Lupus Erythematosus Disease Activity Index; SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index.

  • None of these variables was an independent predictor of indirect costs in the multivariate analysis.

Direct costs     
 Age−0.1520.008   
 Education level0.0620.277   
 Disease duration−0.212< 0.0005−0.013< 0.0005 
 SELENA–SLEDAI score0.1560.006  0.280
 SDI score0.1220.0330.114< 0.0005 
 Number of flares0.448< 0.00010.340< 0.0001 
Indirect costs     
 Age−0.0270.643   
 Education level01   
 Disease duration−0.0200.734   
 SELENA–SLEDAI score0.1020.076   
 SDI score0.0580.308   
 Number of flares0.1360.017   

Severity, organ involvement, and manifestations of flares and costs.

Patients with 1 flare were then grouped into 2 groups: those with a mild/moderate flare (n = 18) and those with a severe flare (n = 32). Patients with mild/moderate and severe flares incurred significantly higher direct costs compared with those without flares (Figure 1A). Patients with severe flares also incurred higher indirect costs compared with those without flares. However, direct and indirect costs did not differ between patients with mild/moderate and severe flares (P = 0.082 for direct costs and P = 0.099 for indirect costs).

thumbnail image

Figure 1. Annual direct and indirect costs by A, severity, B, organ involvement, and C, manifestations of flares. Hatched bars are the indirect costs; stippled bars are the direct costs. * P < 0.005; ** P < 0.0005; # P = 0.044 (P < 0.01 was considered significant after the Bonferroni adjustment). USD = US dollars; NP = neuropsychiatric.

Download figure to PowerPoint

Patients with multiorgan flares incurred significantly higher direct costs compared with those without flares and those with single-organ flares (Figure 1B). Their indirect costs were also higher than those for patients without flares, but this became insignificant after correction for multiple comparisons (P = 0.044).

Patients with single-organ flares were then divided into 2 groups: those with renal/NP flares (n = 27) and those with other manifestations (n = 27). Patients with renal/NP flares generated higher direct costs than those with other manifestations and those without flares (Figure 1C). However, indirect costs did not differ among these 3 groups.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

To our knowledge, this is the first study to elucidate the impact of flares on the costs of SLE. We have shown that patients with flares use more health care resources and incurred both higher direct and indirect costs compared with their counterparts. They paid more visits to health care providers and the emergency room, had a higher hospitalization rate, required more diagnostic examinations, and received more corticosteroids and immunosuppressants. After being adjusted for other sociodeomographics and disease characteristics, the number of flares was an independent explanatory variable associated with increased direct costs in SLE.

There is no generally accepted definition of lupus flare at present, although various approaches have been used in clinical trials (5–7). In our study, we defined flare by encompassing 1) the SELENA–SLEDAI score, 2) clinical disease activity scenarios that may not be captured by the SELENA–SLEDAI descriptors, and 3) change in treatments. Therefore, this should be a comprehensive definition of flare.

The overall flare rate of our cohort was lower compared with previous studies (4, 8, 30). Although this could be due to differences in the definition used, there are several other possible explanations. First, the validity of the retrospective assessment of flare at a specific visit has been shown to be poor (31). Second, the study period in our study is shorter (1 year). Third, although all of the patients were followed at our hospital at regular intervals, it is possible that patients would not seek medical consultation for some minor and short-term flares. Furthermore, we excluded flares with only active serologic manifestations. Although one study has shown the high probability of flares in the next 5 years for patients with serologically active clinically quiescent disease (32), it is our routine practice that we do not launch treatment for these patients. Therefore, it is appropriate to exclude these flares from our analysis.

Sutcliffe et al reported that greater disease activity measured by the Systemic Lupus Activity Measure was associated with high direct and indirect costs (18). A recent study also found that disease activity measured by the Systemic Lupus Activity Questionnaire was an independent predictor of direct health care and productivity costs (17). In our study, the SELENA–SLEDAI was associated with direct costs in univariate analyses, but after being adjusted by other covariates, it became insignificant. Although this may be due to different measures, it may also be explained by the chronic and fluctuating course of SLE that makes an activity score at a single time point not a good indicator of the overall disease activity (16). Therefore, calculating disease activity over time may be desirable (33). The adjusted mean SLEDAI 2000 update (SLEDAI-2K), determined by the calculation of the area under the curve of the SLEDAI-2K over time, has been shown to be strongly associated with mortality (34). Future studies may use this measure to investigate whether disease activity over time is a stronger predictor of costs than disease activity at a single time point.

Since flare is defined as an increase in disease activity, we consider the total number of flares during the study period as a summary of the overall disease activity. In our study, the number of flares was significantly associated with the SELENA–SLEDAI score (r = 0.219, P < 0.0005). The number of flares was significantly related to direct costs, both in the univariate and multivariate models, and it is the only variable significantly associated with indirect costs in the univariate analysis.

In our study, severe flares did not incur significantly higher direct/indirect costs compared with mild/moderate flares. However, we might underestimate the number of mild/moderate flares because a patient might not seek medical consultation for a minor and short-term flare. Multiorgan flares in our study were more costly than single-organ flares. However, it must be noted that the number of patients with multiorgan flares is relatively small. Such a comparison may be of limited value. Flares involving major organs require more aggressive and intensive treatment (12, 35), which concurs with our results that patients with renal/NP flares incur higher direct costs compared with other organ flares. There was no difference in direct costs between patients with renal and NP flares, which is probably due to the small number of patients with NP flares (n = 4).

There are several limitations to this study. The retrospective design may result in inaccuracies of the data, especially for the long recall period (12 months). However, the largest part of the disease costs, i.e., the government health care resources, was obtained by chart review that was solid and accurate. Our indirect costs did not capture the productivity loss because of the time spent nursing patients. However, we included productivity loss in nonpaid work such as housework and daily activity, which are of great importance (36). Although we have shown that patients with flares received more medications, we could not tell from our results whether these medications were initiated during the preceding 12 months or were prescribed because of the flares. Furthermore, because of differences in the patients' sociodemographics, disease features, treatment practices, and health care systems, our results may not be generalizable to other populations of SLE.

In summary, we have shown that patients with flares use more health care resources and incur higher direct and indirect costs compared with those without flares. The total number of flares is an independent explanatory variable associated with direct costs of SLE. Major organ flares such as renal and NP flares incur higher disease costs than other organ flares. Therapies that can effectively control disease activity and prevent flares, especially those that could prevent renal or NP flares, may be cost-effective in view of the high costs associated with active disease affecting these organs. Our results provide some preliminary data for the economic evaluation of such therapies in the future.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Li had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Zhu, Tam, Vivian W.-Y. Lee, Kenneth K.-C. Lee, Li.

Acquisition of data. Zhu, Li.

Analysis and interpretation of data. Zhu, Tam, Li.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

We would like to acknowledge all of the SLE patients for their considerable time and effort contributed toward this project. We also thank our research assistants, Lorraine Tseung and Ka Ling Heung, for their contributions in data collection and entry.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES
  • 1
    Stoll T, Stucki G, Malik J, Pyke S, Isenberg DA. Further validation of the BILAG disease activity index in patients with systemic lupus erythematosus. Ann Rheum Dis 1996; 55: 75660.
  • 2
    Buyon JP, Petri MA, Kim MY, Kalunian KC, Grossman J, Hahn BH, et al. The effect of combined estrogen and progesterone hormone replacement therapy on disease activity in systemic lupus erythematosus: a randomized trial. Ann Intern Med 2005; 142: 95362.
  • 3
    Petri M. Disease activity assessment in SLE: do we have the right instruments? Ann Rheum Dis 2007; 66 Suppl 3: iii614.
  • 4
    Petri M, Genovese M, Engle E, Hochberg M. Definition, incidence, and clinical description of flare in systemic lupus erythematosus: a prospective cohort study. Arthritis Rheum 1991; 34: 93744.
  • 5
    Gladman DD, Urowitz MB, Kagal A, Hallett D. Accurately describing changes in disease activity in systemic lupus erythematosus. J Rheumatol 2000; 27: 3779.
  • 6
    Brunner HI, Feldman BM, Bombardier C, Silverman ED. Sensitivity of the Systemic Lupus Erythematosus Disease Activity Index, British Isles Lupus Assessment Group Index, and Systemic Lupus Activity Measure in the evaluation of clinical change in childhood-onset systemic lupus erythematosus. Arthritis Rheum 1999; 42: 135460.
  • 7
    Buyon JP, Tamerius J, Belmont HM, Abramson SB. Assessment of disease activity and impending flare in patients with systemic lupus erythematosus: comparison of the use of complement split products and conventional measurements of complement. Arthritis Rheum 1992; 35: 102837.
  • 8
    Ho A, Magder LS, Barr SG, Petri M. Decreases in anti–double-stranded DNA levels are associated with concurrent flares in patients with systemic lupus erythematosus. Arthritis Rheum 2001; 44: 23429.
  • 9
    Edwards CJ, Lian TY, Badsha H, Teh CL, Arden N, Chng HH. Hospitalization of individuals with systemic lupus erythematosus: characteristics and predictors of outcome. Lupus 2003; 12: 6726.
  • 10
    D'Cruz DP, Khamashta MA, Hughes GR. Systemic lupus erythematosus. Lancet 2007; 369: 58796.
  • 11
    Stoll T, Sutcliffe N, Mach J, Klaghofer R, Isenberg DA. Analysis of the relationship between disease activity and damage in patients with systemic lupus erythematosus: a 5-yr prospective study. Rheumatology (Oxford) 2004; 43: 103944.
  • 12
    Moroni G, Quaglini S, Gallelli B, Banfi G, Messa P, Ponticelli C. The long-term outcome of 93 patients with proliferative lupus nephritis. Nephrol Dial Transplant 2007; 22: 25319.
  • 13
    Ward MM, Pyun E, Studenski S. Mortality risks associated with specific clinical manifestations of systemic lupus erythematosus. Arch Intern Med 1996; 156: 133744.
  • 14
    Hanly JG, McCurdy G, Fougere L, Douglas JA, Thompson K. Neuropsychiatric events in systemic lupus erythematosus: attribution and clinical significance. J Rheumatol 2004; 31: 215662.
  • 15
    Huscher D, Merkesdal S, Thiele K, Zeidler H, Schneider M, Zink A. Cost of illness in rheumatoid arthritis, ankylosing spondylitis, psoriatic arthritis and systemic lupus erythematosus in Germany. Ann Rheum Dis 2006; 65: 117583.
  • 16
    Lacaille D, Clarke AE, Bloch DA, Danoff D, Esdaile JM. The impact of disease activity, treatment and disease severity on short-term costs of systemic lupus erythematosus. J Rheumatol 1994; 21: 44853.
  • 17
    Panopalis P, Yazdany J, Gillis JZ, Julian L, Trupin L, Hersh AO, et al. Health care costs and costs associated with changes in work productivity among persons with systemic lupus erythematosus. Arthritis Rheum 2008; 59: 178895.
  • 18
    Sutcliffe N, Clarke AE, Taylor R, Frost C, Isenberg DA. Total costs and predictors of costs in patients with systemic lupus erythematosus. Rheumatology (Oxford) 2001; 40: 3747.
  • 19
    Hochberg MC, for the Diagnostic and Therapeutic Criteria Committee of the American College of Rheumatology. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus [letter]. Arthritis Rheum 1997; 40: 1725.
  • 20
    Bombardier C, Gladman DD, Urowitz MB, Caron D, Chang CH, and the Committee on Prognosis Studies in SLE. Derivation of the SLEDAI: a disease activity index for lupus patients. Arthritis Rheum 1992; 35: 63040.
  • 21
    Gladman D, Ginzler E, Goldsmith C, Fortin P, Liang M, Urowitz M, et al. The development and initial validation of the Systemic Lupus International Collaborating Clinics/American College of Rheumatology damage index for systemic lupus erythematosus. Arthritis Rheum 1996; 39: 3639.
  • 22
    Zhu TY, Tam LS, Lee VW, Hwang WW, Li TK, Lee KK, et al. Costs and quality of life of patients with ankylosing spondylitis in Hong Kong. Rheumatology (Oxford) 2008; 47: 14225.
  • 23
    Linnik MD, Hu JZ, Heilbrunn KR, Strand V, Hurley FL, Joh T, and the LJP 394 Investigator Consortium. Relationship between anti–double-stranded DNA antibodies and exacerbation of renal disease in patients with systemic lupus erythematosus. Arthritis Rheum 2005; 52: 112937.
  • 24
    Alarcon-Segovia D, Tumlin JA, Furie RA, McKay JD, Cardiel MH, Strand V, et al, and the LJP 394 Investigator Consortium. LJP 394 for the prevention of renal flare in patients with systemic lupus erythematosus: results from a randomized, double-blind, placebo-controlled study. Arthritis Rheum 2003; 48: 44254.
  • 25
    ACR Ad Hoc Committee on Neuropsychiatric Lupus Nomenclature. The American College of Rheumatology nomenclature and case definitions for neuropsychiatric lupus syndromes. Arthritis Rheum 1999; 42: 599608.
  • 26
    Leung GM, Bacon-Shone J. Hong Kong's health system: reflections, perspectives and visions. Hong Kong: Hong Kong University Press; 2006.
  • 27
    Grant C. The Hong Kong health care systems. Sydney: University of New South Wales; 1998.
  • 28
    Woo J, Lau E, Lau CS, Lee P, Zhang J, Kwok T, et al. Socioeconomic impact of osteoarthritis in Hong Kong: utilization of health and social services, and direct and indirect costs. Arthritis Rheum 2003; 49: 52634.
  • 29
    Liljas B. How to calculate indirect costs in economic evaluations. Pharmacoeconomics 1998; 13: 17.
  • 30
    Weiss JE, Sison CP, Ilowite NT, Gottlieb BS, Eberhard BA. Flares in pediatric systemic lupus erythematosus. J Rheumatol 2007; 34: 13414.
  • 31
    FitzGerald JD, Grossman JM. Validity and reliability of retrospective assessment of disease activity and flare in observational cohorts of lupus patients. Lupus 1999; 8: 63844.
  • 32
    Ng KP, Manson JJ, Rahman A, Isenberg DA. Association of antinucleosome antibodies with disease flare in serologically active clinically quiescent patients with systemic lupus erythematosus. Arthritis Rheum 2006; 55: 9004.
  • 33
    Nossent JC. Course and prognostic value of Systemic Lupus Erythematosus Disease Activity Index in black Caribbean patients. Semin Arthritis Rheum 1993; 23: 1621.
  • 34
    Ibanez D, Urowitz MB, Gladman DD. Summarizing disease features over time. I. Adjusted mean SLEDAI derivation and application to an index of disease activity in lupus. J Rheumatol 2003; 30: 197782.
  • 35
    Hermosillo-Romo D, Brey RL. Diagnosis and management of patients with neuropsychiatric systemic lupus erythematosus (NPSLE). Best Pract Res Clin Rheumatol 2002; 16: 22944.
  • 36
    Clarke AE, Penrod J, St Pierre Y, Petri MA, Manzi S, Isenberg DA, et al, and the Tri-Nation Study Group. Underestimating the value of women: assessing the indirect costs of women with systemic lupus erythematosus. J Rheumatol 2000; 27: 2597604.