Abstract
The cause of death in people affected by sickle cell disease (SCD) is often challenging to define as prior studies have used retrospective or administrative data for analysis. We used a prospective longitudinal registry to assess mortality and clinical co-morbidities among subjects enrolled in the Sickle Cell Disease Implementation Consortium (SCDIC) registry. At enrollment, we collected the following data: patient-reported demographics, SCD phenotype, baseline laboratory values, comorbidities, and current medications. Subjects were followed for a median of 4.7 years before the present analysis. The relationship of clinical co-morbidities (at time of enrollment) to mortality was determined using survival analysis, adjusting for SCD phenotype and gender. There was a total of 2,439 people with SCD enrolled in the SCDIC registry. One hundred and twenty-eight participants (5%) died during the observation period (2017-2022). Six people died from trauma and were excluded from further analysis. Proximate cause of death was unwitnessed in 17% of the deaths, but commonest causes of death include cardiac (18%), acute chest or respiratory failure (11%), sudden unexplained death (8%). Enrollment characteristics of the individuals who died (n=122) were compared to those of survivors (n=2317). Several co-morbidities at enrollment increased the odds of death on univariate analysis. All co-morbidities were included in a multivariable model. After backward elimination, iron overload, pulmonary hypertension, and depression, remained statistically significant predictors of the risk of death. SCD reduces life expectancy. Improved comprehensive and supportive care to prevent end-organ damage and address comorbidities is needed for this population.
Keywords: Sickle cell disease, SCD Mortality, SCD survival, SCDIC
Introduction.
Sickle cell disease (SCD) results from the inheritance of two mutated β-globin gene (one from each parent). SCD is a broad term describing a group of inherited multisystemic diseases including Hb SS or sickle cell anemia (SCA), HbSC, and HbSβ-thalassemia. These conditions are characterized by episodes of acute illness and progressive organ damage.(1) SCD is a common childhood-onset genetic disorder mainly affecting persons of African American origin in the United States. (2) The pathophysiological manifestation of this single-gene disorder is characterized by hemoglobin polymerization, which induces red cell deformation in various conditions of deoxygenation. It is responsible for chronic hemolytic anemia, acute microvascular occlusions, tissue infarctions, and endothelial dysfunction. (1, 3–5) Over time, chronic multi-organ damage becomes cumulative, increasing the risk of early death.(6)
During the past few decades, the probability that children with SCD in high-resource countries will survive into adulthood has increased due to effective clinical interventions including newborn screening, clinical education, penicillin prophylaxis, pneumococcal vaccination, hydroxyurea therapy, improved recognition and treatment of splenic sequestration, and blood transfusion. (7–10) However, SCD-induced organ damage and reduced access to healthcare resources (specifically to SCD specialists), continue to challenge the survival of affected young adults. Several studies have demonstrated that mortality in adults with SCD has not improved significantly despite significant childhood improvements. (11–13) Although research has shown that SCD results in chronic multi-organ complications, it is hard to predict which individuals are at risk for these specific complications or for early mortality. SCD prognosis has previously been defined partly by SCD genotype but also by clinical phenotypes that span across genotypes; however this has not been prospectively assessed in a large modern, natural history study.(14, 15) In a 2017 meta-analysis by Maitra et.al, age, increased tricuspid regurgitant jet velocity, lower baseline hemoglobin, increased creatinine, and increased NT-pro-BNP were associated with increased risk of mortality. (16) Other investigators have reported that end stage renal disease, co-morbid seizure disorder, history of acute chest syndrome (ACS), low fetal hemoglobin (Hb) level, and baseline elevated white blood cell (WBC) count greater than 15,000 cells per cubic millimeter were associated with decreased survival. (17) Despite these reports, it is difficult to predict which children with SCD are at high risk of early mortality and may require more intensive SCD-modifying therapy. Further, the cause of death in adults with SCD is often challenging to define due to retrospective or administrative data-based analysis or the more general terms used in death certificates (heart failure, respiratory failure) that do not aptly describe what led to the actual death. (18)Thus, prospective longitudinal clinical registries are needed to enhance understanding of the cause of death and identify strategies for early intervention. Understanding the cause of mortality in SCD also provides information relevant to clinical care, patient-oriented research, therapy development, public health planning, and patient, family, and provider education.(18)
The current study aims to describe the factors that portend positive and negative prognoses in SCD and the clinical events leading to death among adolescents and adults in the United States during the years 2017-2022 using data from the National Institution of Health funded Sickle Cell Disease Implementation Consortium (SCDIC) registry. The SCDIC was a cooperative research program that included eight academic hospital clinical sites spread across the mainland United States and a data coordinating center. (19) We hypothesized that specific co-morbidities or baseline laboratory findings would predict the risk of death in this population. Furthermore, we also hypothesized that while the mortality rates would continue to differ between SCD genotypes (sickle cell anemias (SCA)-HbSS and HBSB0 thalassemia and non-sickle cell anemias (non-SCA) including HbSC and HbSβ+ thalassemia), the causes of mortality would be similar.
Methods
Study Population.
The study population included adolescent and adult patients from eight SCD treatment centers across the United States enrolled in the SCDIC Registry. (19, 20) Participants in the registry were eligible for recruitment based on the following inclusion criteria: 15 to 45 years of age, confirmed diagnosis of SCD (hemoglobin, Hb SS, Hb SC disease, Hb Sβ0 thalassemia, Hb Sβ+ thalassemia, Hb S hereditary persistence of fetal Hb (S/HPFH), Hb SE, Hb SD, or Hb SO Arab), literacy in English, and willingness to provide informed consent or assent. Participants were excluded from enrollment if they had sickle cell trait (hemoglobin AS) or had undergone a successful hematopoietic stem cell transplant. The protocol stated that participants should be approached during routine outpatient visits and while they were in steady state, but this was not a strict requirement. As previously described, the protocol called for a baseline visit and two annual follow-up visits. (21). Baseline visits took place in 2016-2019. This study was approved by the institutional review boards at all sites and written informed consent was obtained from all participants or their legal guardian.
Data Collection and Measures.
Data for the SCDIC registry were collected from the medical record and by use of participant self-report surveys. Additionally, the primary investigator at each site was asked to review pertinent information from the medical records and any other available sources (such as death certificates, autopsy reports, and family members) to provide any further insights into the cause(s) of each participant death, as well as the date each participant was last known to be alive. From the participant self-report surveys, we collected socio-demographic measures including age, gender, race, ethnicity, primary language, marital status, number of children/adults living in the household, income, education, and employment status. Other patient-reported outcomes collected include frequency of pain medication for SCD, frequency/timing of pain attacks, any past/current depression treatment, absence/delay of medical care, and current difficulties with cognitive tasks and task management. From the medical records, we collected the participants’ SCD phenotype, blood transfusion history, number of hospital admissions in the past year, history of SCD complications, other co-morbid conditions, any cancer diagnoses, and most recent standard laboratory measures. Imaging data was also captured if available for any participant with pulmonary hypertension (PH) or left ventricular (LV) dysfunction indicated on the baseline medical record abstraction form. The diagnosis of PH and LV dysfunction were not standardized for this project and were only obtained from medical history.
Outcome of Interest.
The primary focus of our analysis was understanding the causes of death for people ages 15-45 year and the associated risk factors. Details regarding each patient’s death was captured from all/any available sources (death certificates, medical records, autopsy reports, and/or reports from family members); and identified both primary and secondary causes of death in consultation with the primary investigators. The results were then categorized as (1) multi-organ failure, (2) acute chest / pulmonary, (3) cardiac, (4) sudden death, (5) leukemia / myelodysplastic syndrome, (6) other cancer, (7) unknown, (8) trauma, (9) liver failure, (10) kidney failure, (11) stroke, (12) COVID-19, (13) Sepsis, (14) transplant failure, (15) hemorrhagic shock, (16) suicide, (17) drug overdose, or (18) pulmonary embolism. Where appropriate, participants were assigned to more than one category. There were six people in the registry who died because of trauma who were removed prior to analysis.
Statistical Analysis.
Baseline characteristics of the study population were summarized overall and by death status. The study population was also summarized by non-SCA (Hb SC disease, Hb Sβ+ thalassemia, Hb S hereditary persistence of fetal Hb (S/HPFH)) and SCA (Hb SS, Hb Sβ0 thalassemia, Hb SE, Hb SD, Hb SO) phenotype. The summaries included frequency and percentage for categorical variables and median, interquartile ranges for continuous variables. Differences in characteristics were determined by Chi-square, Fisher’s exact, Wilcoxon rank sum as appropriate. The effects of comorbidities on risk of death were examined using proportional hazards regression with age as the time metric. All models included gender and SCD phenotype as clinically relevant covariates. First, each comorbidity under consideration was included in a separate model. Then, backward elimination was applied to a multivariable model to exclude all comorbidities that were not statistically significant predictors. A significance cutoff of 0.05 was used for backward elimination. Follow-up of participants who did not die during the study was right censored at the age they were last known to be alive. Follow-up was also left truncated at the age of study entry. We included gender, genotype, and all SCD complications into the initial model. Results included the model-based hazard ratio estimates and p-values. Reference groups for hazard ratios were female, non-SCA, and those without the SCD complication or condition, with hazard ratios < 1.0 indicating a smaller risk of death and hazard ratios > 1.0 indicating increased risk of death. Analyses were conducted in SAS Version 9.4 (SAS Institute Inc., Cary, NC, USA).
Results
Baseline characteristics.
A total of 2,439 registry participants were followed for a median of 4.68 years. Baseline characteristics of those who died (n=122, 5%) and those who did not die during the study (n=2,317, 95%) are presented in Table 1a, b. Age at enrollment was divided into several cohorts (<18 years, 18-24 years, 25-34 years, and 35-45 years). We stratified the 122 deceased participants by genotype (SCA: n=107, 88%; non-SCA: n=15, 12%) in Table 1b. Mortality rates increased with ascending age group cohort from 97 per 100,000 person years in the <18 years cohort to 726, 1247, 1553 per 100, 000 person years in the 18-24, 25-34, 35-45 years age groups respectively. Table 1a.
Table 1a.
Demographic characteristics of deceased vs living participants at the end of the observation period.
| Characteristic | Deaths (N=122) | Alive (N=2317) | Total (N=2439) |
|---|---|---|---|
| Age group (years) | |||
| <18 | 1 (0.8%) | 220 (9.5%) | 221 (9.1%) |
| 18-24 | 23 (18.9%) | 654 (28.2%) | 677 (27.8%) |
| 25-34 | 57 (46.7%) | 920 (39.7%) | 977 (40.1%) |
| 35-45 | 41 (33.6%) | 523 (22.6%) | 564 (23.1%) |
| Age (years) | |||
| Median | 32.0 | 27.0 | 27.0 |
| Gender | |||
| Male | 47 (38.5%) | 1,004 (43.3%) | 1,051 (43.1%) |
| Female | 75 (61.5%) | 1,313 (56.7%) | 1,388 (56.9%) |
| Genotype | |||
| Hb SS | 105 (86.1%) | 1,582 (68.4%) | 1,687 (69.3%) |
| Hb SC | 13 (10.7%) | 486 (21.0%) | 499 (20.5%) |
| Hb S beta0 | 1 (0.8%) | 91 (3.9%) | 92 (3.8%) |
| Hb S beta+ | 2 (1.6%) | 130 (5.6%) | 132 (5.4%) |
| Hb S/HPFH | 0 (0.0%) | 14 (0.6%) | 14 (0.6%) |
| Hb SE | 0 (0.0%) | 1 (0.0%) | 1 (0.0%) |
| Hb SD | 0 (0.0%) | 4 (0.2%) | 4 (0.2%) |
| Hb SO | 1 (0.8%) | 5 (0.2%) | 6 (0.2%) |
| Race | |||
| Black or African American | 119 (97.5%) | 2,210 (96.9%) | 2,329 (97.0%) |
| Other | 3 (2.5%) | 70 (3.1%) | 73 (3.0%) |
| Marital Status | |||
| Married or living as married | 10 (8.7%) | 327 (16.4%) | 337 (15.9%) |
| Other | 105 (91.3%) | 1,671 (83.6%) | 1,776 (84.1%) |
| Household income | |||
| $25,000 or less | 70 (70.7%) | 1,075 (53.1%) | 1,145 (53.9%) |
| $25,001+ | 29 (29.3%) | 949 (46.9%) | 978 (46.1%) |
| Education | |||
| Less than high school graduate | 9 (7.8%) | 403 (17.9%) | 412 (17.4%) |
| High school graduate or higher | 106 (92.2%) | 1,854 (82.1%) | 1,960 (82.6%) |
| Employment | |||
| Engaged | 74 (64.3%) | 899 (40.2%) | 973 (41.3%) |
| Unengaged | 41 (35.7%) | 1,340 (59.8%) | 1,381 (58.7%) |
Table 1b.
Mortality categorized by SCD phenotype and gender. Sickle cell anemias = Hb SS, Hb S beta0, Hb SE, Hb SD, Hb SO. non sickle cell anemias = Hb SC, Hb S beta+, Hb S/HPFH
| Genotype and Sex | No of Patients | Age at entry (Yr) Median (IQR) | follow-up (Yr) | No of deaths |
|---|---|---|---|---|
| Sickle cell anemia | 1790 | 28 (22, 34) | 4.70 (4.35, 5.01) | 107 |
| Male | 788 | 27 (21, 33) | 4.71 (4.36, 5.01) | 41 |
| Female | 1002 | 28 (22, 34) | 4.70 (4.33, 5.01) | 66 |
| Non sickle cell anemia | 645 | 27 (22, 34) | 4.66 (4.35, 4.97) | 15 |
| Male | 263 | 27 (21, 34) | 4.68 (4.35, 4.97) | 6 |
| Female | 382 | 28 (22, 34) | 4.64 (4.35, 4.97) | 9 |
Risk factors affecting age at death.
When the comorbidities were examined in separate proportional hazards models, risk of death was found to be statistically significant for participants with pulmonary hypertension, end stage renal disease, deep vein thrombosis, sarcoidosis, diabetes mellitus, iron overload, chronic refractory pain, or depression (Table 2). Among the conditions presented in Table 2, it is also noted that having two or more of these complications further increased the association with early mortality (Figure 1b). However, there is little we can say about the risks associated with having multiple comorbidities as we did not explore the interaction of the individual comorbid conditions or the actual mix of complications. For instance, a patient with two complications, neither of which is strongly linked to risk of death is likely to be very different from a patient with two complications that are both linked to elevated risk of death. In univariate analysis, many of the conditions in Table 2 showed significant association with mortality, but this association was attenuated in multivariable analysis when accounting for SCD phenotype, gender, and collinearity with other conditions, reflecting the strong association between presence of SCA and organ damage. However, iron overload, depression and pulmonary hypertension retained their independent association with mortality. (Figure 1c) Despite the association of SCA with clinical severity and mortality, we note that many individuals with HbSC disease also had a significant disease burden; people with HbSC who died had a history of joint necrosis (40%), CKD (20%), pulmonary hypertension (27%), DVT (33%), chronic refractory pain (27%), and depression (47%).
Table 2.
Co-morbid conditions categorized by mortality status. P-values are from individual Cox Proportional Hazard models, including gender and genotype, left-truncated for age at study entry. “Yes” column shows the percentage of participants with a diagnosed comorbid condition, at enrollment, who died during the observation period. “No” column shows the percentages of participants without a diagnosis of a similar comorbid condition, at enrollment, who died during the observation period.
| Comorbidity at enrollment | Yes | No | ||
|---|---|---|---|---|
| % dead (Total) | % dead (Total) | Total | P-Value | |
| Avascular necrosis | 8.1 (632) | 4.4 (1079) | 1711 | 0.1068 |
| Dactylitis | 6.0 (100) | 5.8 (1020) | 1120 | 0.5975 |
| Osteomyelitis | 8.2 (110) | 6.0 (1143) | 1253 | 0.9001 |
| Chronic kidney disease | 11.6 (198) | 5.5 (1245) | 1443 | 0.0624 |
| End stage renal disease | 27.3 (22) | 5.9 (1222) | 1244 | 0.0036 |
| Priapism | 4.9 (247) | 5.4 (1490) | 1737 | 0.5072 |
| Stroke | 7.8 (384) | 4.6 (1351) | 1735 | 0.1584 |
| Intracranial bleeding | 0.0 (26) | 6.4 (1140) | 1166 | 0.9875 |
| Pulmonary hypertension | 12.9 (286) | 4.7 (1204) | 1490 | <.0001 |
| Left ventricular dysfunction | 10.5 (76) | 5.9 (1209) | 1285 | 0.3755 |
| Acute chest syndrome | 6.0 (1337) | 3.7 (649) | 1986 | 0.0857 |
| Asthma | 5.8 (617) | 5.4 (1024) | 1641 | 0.1818 |
| Gallstones/cholelithiasis, cholecystitis | 6.7 (1072) | 3.7 (728) | 1800 | 0.5073 |
| Splenomegaly: Hypersplenism | 0.0 (61) | 6.0 (1012) | 1073 | 0.1950 |
| Deep vein thrombosis (DVT) | 9.8 (378) | 5.1 (1163) | 1541 | 0.0366 |
| Lupus | 0.0 (11) | 6.5 (1295) | 1306 | 0.9816 |
| Rheumatoid arthritis | 10.5 (19) | 6.4 (1256) | 1275 | 0.6247 |
| Gout | 5.6 (18) | 6.6 (1234) | 1252 | 0.7055 |
| Sarcoidosis | 62.5 (8) | 6.4 (1201) | 1209 | <.0001 |
| Multi-organ failure: all types | 11.2 (116) | 6.3 (1062) | 1178 | 0.1008 |
| Pneumococcal sepsis | 5.4 (92) | 6.0 (1191) | 1283 | 0.5120 |
| Skin ulcers | 9.7 (93) | 6.0 (1293) | 1386 | 0.4680 |
| Retinopathy | 3.1 (324) | 5.7 (1291) | 1615 | 0.0750 |
| Diabetes mellitus (other systemic) | 13.2 (53) | 5.7 (1307) | 1360 | 0.0276 |
| Iron overload (Other) | 11.0 (575) | 3.7 (945) | 1520 | 0.0005 |
| Chronic refractory pain (Other) | 9.6 (489) | 4.2 (884) | 1373 | 0.0070 |
| Anxiety (Mental health) | 8.6 (315) | 5.4 (1191) | 1506 | 0.4816 |
| Depression (Mental health) | 9.1 (464) | 4.6 (1136) | 1600 | 0.0111 |
Figure 1.
Panel a) Causes of death by frequency in the SCDIC cohort. A participant could be assigned to more than one cause of death category. b) Frequency of two or more clinical comorbidities in alive vs dead participants. Participants with two or more end-organ dysfunctions at enrollment showed decrease survival. c) Hazard ratio of death with 95%CI for comorbid conditions, at the time of enrollment. Iron overload, depression and pulmonary hypertension retained their independent association with mortality, in the final model.
Multivariable modeling.
The initial multivariable model included gender, SCD phenotype, and all the comorbidities in Table 2. The Cox proportional hazards model for gender and SCD phenotype showed no increased risk of death by gender (Figure 2a, c, HR= 0.79, p= 0.246), but a significantly increased risk of death by SCD phenotype (Figure 2b, c, HR= 2.44, p=0.003). We retained both covariates in our final model, regardless of statistical significance, due to their clinical importance. After backward elimination, four comorbidities remained in the model with gender and SCD phenotype: iron overload (HR= 3.26, p= 0.001), depression (HR= 3.24, p= 0.0006), pulmonary hypertension (HR= 3.35, p= 0.0045) and gout (HR 17.44, p=0.017). Note that only 18 participants had a diagnosis of gout at enrollment and 1/18 died during observation. Therefore, the association of gout with mortality here should be interpreted with caution because of the small number of affected persons. The diagnosis of pulmonary hypertension was per the medical record and not confirmed by imaging or adjudication.
Figure 2.
Survival of patients in the SCDIC cohort a) Shows male and female patient of all SCD genotypes. The Cox proportional hazards model showed no increased risk of death by gender (HR= 0.79, p= 0.246), but shows b) a significantly increased risk of death by genotype (HR= 2.44, p=0.003) c) shows the impact of genotype and gender on survival by age.
Cause of Death.
The causes of death listed most frequently in the registry were categorized as cardiac (18%), acute chest syndrome (11%), multi-organ failure syndrome (MOFS) (11%), sudden death (8%), sepsis (7%), Liver failure (4%), and kidney failure (4%) (Figure 1a). Cause of death was listed as undetermined in 18% of total recorded deaths. Severe organ damage prior to death was noted in all cases with undetermined cause of death. The cause of death was determined from various sources-medical record (66), family members (14), family member+ medical record (5), medical record+ death certificate (11), death certificate alone (2), death certificate + family member (3), medical record + autopsy (1), autopsy alone (1). Unfortunately, in 25 deaths, no source of information related to the person’s death was available, but the participant was flagged as dead in the state registry or electronic medical record. The PI from each center reviewed each death to determine the cause of death with as much details as possible. Organ damage was caused by multi-organ failure syndrome (MOFS) or severe intrahepatic cholestasis.(21) Most cases of sudden death were cases of collapse at home with witnesses or individuals who were found dead at home. The role of opioid medications in these scenarios remains unclear but was implicated by family members in some reports. Most deaths from pulmonary embolism occurred in concert with acute chest syndrome or cardiac death. Liver and kidney deaths often occurred alongside other organ failures, notably acute chest syndrome or cardiac death. Other less common causes of death included acute stroke (2%), solid organ or stem cell transplant failure (2%), drug overdose (2%), COVID-19 infection (1%), leukemia/MDS (1%), other cancers (2%), and suicide (1%). Isolated trauma was the cause of death in 6 (3%) of the deceased participants.
Discussion
These findings provide an updated understanding of the burden of SCD complications and causes of death in individuals with SCD in the United States between the years 2017 and 2022. Our analysis demonstrates that adults with SCD have a substantial risk of developing multi-organ complications which may increase the risk of mortality. Notably, while there are more people in the registry who died with sickle cell anemia, those with HbSC disease who died also had a high burden of co-morbidities. It is therefore important to recognize the severity of all forms of SCD irrespective of phenotype. (22, 23)
We also interrogated the actual causes of mortality in the deceased cohort, given the paucity of aggregated prospective mortality data in adult patients with SCD. After a total follow-up period of 11,350 person years and a median time of follow-up of 4.7 years for all patients, 122 patients (5%) died. The overall mortality rate for our 15-45-year-old cohort was 1075 deaths per 100,000 person years. This is similar to findings of a retrospective study performed in England that described the mortality rate in an adult SCD cohort of 950/100,000 person years. (24) In our cohort, death from cardiopulmonary causes was the most common. Although prior epidemiological studies have highlighted the importance of cardiopulmonary death in adults with SCD, (25) this is the first study to confirm this observation using a prospective cohort where cause of death was adjudicated individually. In this cohort, the primary cause of death was cardiac in 18% and pulmonary in 11% in contrast to prior epidemiological study that recorded cardiac and pulmonary causes as main cause of death in 31.6% and 28.1% of a cohort of 5,223 deceased SCD patients. (25) We however acknowledge that 8% those who died in our cohort had sudden death that could not be further characterized, and may have been the result of fatal arrhythmias, acute pulmonary hypertensive crisis, drug toxicity, or other unknown causes.
Mortality rate in our cohort increased with increasing age and we observed the highest mortality rate of 1553 per 100,000 person years in our oldest age group, the 35-45 years cohort, during the observation period. This is similar to prior epidemiological data showing that all-cause mortality rate for SCD patient in the states of Georgia and California increased sharply in adolescents and continued to rise with older age group cohorts.(18) Other investigators have found that people with SCD have an increased risk of acute care utilization and pain frequency in late adolescence, as well as an increased risk of mortality. It is plausible that the concomitant socio-demographic challenges previously described in this population, reliance on public health insurance, cognitive challenges, and a high rate of comorbid disease also increased the risk of mortality in this vulnerable period. (26–28) It is also reasonable that people with SCD who utilize acute care more frequently may not be seeing physicians with SCD expertise in the acute care setting.
In our cohort, having a diagnosis of pulmonary hypertension at the time of enrollment raised the risk of dying at any point in time by three-fold, independent of other predictors. Previous studies have shown that pulmonary hypertension in SCD is associated with high mortality and an elevated TRV on echocardiogram is a well-described marker of poor prognosis. (29, 30)
One of the non-trivial recurring co-morbidities associated with increased morbidity and mortality in adults with SCD is a history of depression and or anxiety disorder. Chronically ill patients are at risk for depression, which can affect health-related quality of life, health care utilization, and cost. (31) Previous studies in the general population and SCD population, have linked depression with mortality. (31, 32) For 2019, the National Center for Health Statistics estimates that the prevalence of at least moderate depression in the non-Hispanic black population was about 4.6% (33). The prevalence of depression in our deceased cohort was 29% and this is about six times higher than that of the general black population. Prior studies have demonstrated that improving depression management in patients with chronic medical conditions comes with appreciable cost savings in reduced healthcare utilization and decreases mortality. (34)
This cohort showed high burden of iron overload overall as well as in those who died. Iron overload was identified in the medical record for this study and was not always measured using magnetic resonance imaging (MRI) of the liver. (35) In support of this finding, we observed that SCD subjects who died had a median ferritin level >5 times the level of the alive cohort consistent with the reports of iron overload in the medical record. Our findings are supported by other studies noting that iron overload is associated with higher incidence of acute painful episodes, organ failure and predicts mortality in SCD patients. (36, 37) Iron overload is caused by red cell transfusions in SCD. In some cases, the need for frequent transfusions could signal more severe SCD; (38) however, in other cases, people may have received their care in community hospitals where SCD specialists were not available to limit the amount of transfusions given during uncomplicated pain crisis.
No significant association was observed between the use of hydroxyurea and mortality in our analysis. However, there is insufficient data on medication adherence to hydroxyurea, duration of therapy, age at initiation of therapy or hydroxyurea doses available in the study. New therapeutic options such as L-glutamine, crizanlizumab, and voxelotor were not uniformly available and were not included in this study. Our study did not demonstrate any independent association between most baseline laboratory data and mortality.
This study has several limitations. The primary limitation is the age of enrollment in the registry (limited to ages 15-45 years). Future studies on adult morbidity and mortality should include all age groups for broader representation, including those older than 50, who are often underrepresented in clinical trials and qualitative studies on SCD. Further, the analysis of co-morbidities relied on the completeness and accuracy of the medical record and patient self-report. Ideally, all co-morbidities included in the SCDIC registry would have met a standard definition. Instead, co-morbidities in this study were included if they were noted in the medical record. Further, there are numerous instances in which information on a given condition/comorbidity was listed as “not present in the medical record”. Not present in the record is obviously not the same as is absent; thus, patients in the registry could have had additional complications that were not included in the registry. There were also challenges in accurately assessing the cause of death due to unavailable or incomplete records. Finally, all the SCDIC sites included in the registry are academic sites with well-established SCD centers. Thus, despite participation from multiple locations across the U.S., the generalizability of the sample may still be limited and may not be representative of SCD individuals disconnected from care.
These data support the importance of preventative care in SCD, especially for monitoring and surveillance of organ system function in affected adults so that early treatment can be initiated. Further, the complexity of the co-morbidities in SCD continues to highlight the need for all individuals to see an SCD-specialist in collaboration with a primary care provider. SCD patients with progressive end-organ disease should be considered for referral to a specialist in the disease domain for longitudinal monitoring. Assessments of depression and anxiety should be performed in all persons living with SCD and appropriate treatment should be available, affordable, and accessible. In multiple states, the limited reimbursement by Medicaid for services such as psychological therapy or psychiatric assessment highly limit their accessibility. (39) Finally, all individuals with SCD need comprehensive care, regardless of genotype which will require novel methods of access to care and improved payer mechanisms to ensure sufficient treatment. As individuals with SCD grow older, the need for ongoing preventative care is imperative to improving outcomes and longevity. Additional federal funding for a national, longitudinal, clinical registry and greater collaboration and alignment across SCD centers is needed to better predict and prevent early mortality in SCD.
Funding
The Sickle Cell Disease Implementation Consortium (SCDIC) has been supported by US Federal Government cooperative agreements HL133948, HL133964, HL133990, HL133996, HL133994, HL133997, HL134004, HL134007, and HL134042 from the National Heart Lung and Blood Institute and the National Institute on Minority Health and Health Disparities (Bethesda, MD).
Declarations Conflict of interest
NS receives research funding from Global Blood Therapeutics, is a consultant for Novartis, Emmaus Medical and Forma Therapeutics, Agios and a speaker for Global Blood Therapeutics and Novartis. VRG has served as a consultant for CSL Behring.
Footnotes
Ethical approval
This study was approved by the institutional review boards of the participating institutions (Approval Numbers: University of California San Francisco 20–31161; Research Triangle Institute 14157; Washington University, St. Louis 201706016; Mount Sinai PPHS 16–01047; University of Illinois, Chicago 17–061401; Duke University Pro00073506; Augusta University 1078191). The authors certify that all procedures in the study were performed in accordance with the ethical standards as laid down in the 1964 declaration of Helsinki and its latter amendments or comparable ethical standards.
Informed consent
Informed consent to participate in the study was obtained from all participants or their legal guardian, if participant was a minor.
Consent for publication
Informed consent included agreement to have de-identified data published and disseminated in accordance with publication guidelines for the Sickle Cell Disease Implementation Consortium.
Data Availability Statement
The data that support the findings of this study are available from Department of Clinical Research, SSES, RTI International, Research Triangle Park, North Carolina, USA, upon reasonable request.
References:
- 1.Bunn HF. Pathogenesis and Treatment of Sickle Cell Disease. New England Journal of Medicine. 1997;337(11):762–9. [DOI] [PubMed] [Google Scholar]
- 2.Hassell KL. Population estimates of sickle cell disease in the U.S. Am J Prev Med. 2010;38(4 Suppl):S512–21. [DOI] [PubMed] [Google Scholar]
- 3.Brittenham GM, Schechter AN, Noguchi CT. Hemoglobin S polymerization: primary determinant of the hemolytic and clinical severity of the sickling syndromes. Blood. 1985;65(1):183–9. [PubMed] [Google Scholar]
- 4.Noguchi CT, Schechter AN. Non-uniformity of intracellular polymer formation in sickle erythrocytes: possible correlation with severity of hemolytic anemia. Am J Pediatr Hematol Oncol. 1984;6(1):46–50. [PubMed] [Google Scholar]
- 5.Klug PP, Lessin LS, Radice P. Rheological Aspects of Sickle Cell Disease. Archives of Internal Medicine. 1974;133(4):577–90. [PubMed] [Google Scholar]
- 6.Chaturvedi S, Ghafuri DL, Jordan N, Kassim A, Rodeghier M, DeBaun MR. Clustering of end-organ disease and earlier mortality in adults with sickle cell disease: A retrospective-prospective cohort study. Am J Hematol. 2018;93(9):1153–60. [DOI] [PubMed] [Google Scholar]
- 7.Gaston MH, Verter JI, Woods G, Pegelow C, Kelleher J, Presbury G, et al. Prophylaxis with Oral Penicillin in Children with Sickle Cell Anemia. New England Journal of Medicine. 1986;314(25):1593–9. [DOI] [PubMed] [Google Scholar]
- 8.Wang WC, Ware RE, Miller ST, Iyer RV, Casella JF, Minniti CP, et al. Hydroxycarbamide in very young children with sickle-cell anaemia: a multicentre, randomised, controlled trial (BABY HUG). Lancet. 2011;377(9778):1663–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hyacinth HI, Adams RJ, Voeks JH, Hibbert JM, Gee BE. Frequent red cell transfusions reduced vascular endothelial activation and thrombogenicity in children with sickle cell anemia and high stroke risk. Am J Hematol. 2014;89(1):47–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Halasa NB, Shankar SM, Talbot TR, Arbogast PG, Mitchel EF, Wang WC, et al. Incidence of invasive pneumococcal disease among individuals with sickle cell disease before and after the introduction of the pneumococcal conjugate vaccine. Clin Infect Dis. 2007;44(11):1428–33. [DOI] [PubMed] [Google Scholar]
- 11.Payne AB, Mehal JM, Chapman C, Haberling DL, Richardson LC, Bean CJ, et al. Trends in Sickle Cell Disease-Related Mortality in the United States, 1979 to 2017. Ann Emerg Med. 2020;76(3s):S28–s36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Quinn CT, Rogers ZR, McCavit TL, Buchanan GR. Improved survival of children and adolescents with sickle cell disease. Blood. 2010;115(17):3447–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lanzkron S, Carroll CP, Haywood C, Jr. Mortality rates and age at death from sickle cell disease: U.S., 1979-2005. Public Health Rep. 2013;128(2):110–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Steinberg MH, Sebastiani P. Genetic modifiers of sickle cell disease. Am J Hematol. 2012;87(8):795–803. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Quinn CT. Minireview: Clinical severity in sickle cell disease: the challenges of definition and prognostication. Exp Biol Med (Maywood). 2016;241(7):679–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Maitra P, Caughey M, Robinson L, Desai PC, Jones S, Nouraie M, et al. Risk factors for mortality in adult patients with sickle cell disease: a meta-analysis of studies in North America and Europe. Haematologica. 2017;102(4):626–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Platt OS, Brambilla DJ, Rosse WF, Milner PF, Castro O, Steinberg MH, et al. Mortality in sickle cell disease. Life expectancy and risk factors for early death. N Engl J Med. 1994;330(23):1639–44. [DOI] [PubMed] [Google Scholar]
- 18.Paulukonis ST, Eckman JR, Snyder AB, Hagar W, Feuchtbaum LB, Zhou M, et al. Defining Sickle Cell Disease Mortality Using a Population-Based Surveillance System, 2004 through 2008. Public Health Rep. 2016;131(2):367–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.DiMartino LD, Baumann AA, Hsu LL, Kanter J, Gordeuk VR, Glassberg J, et al. The sickle cell disease implementation consortium: Translating evidence-based guidelines into practice for sickle cell disease. Am J Hematol. 2018;93(12):E391–E5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Knisely MR, Pugh N, Kroner B, Masese R, Gordeuk V, King AA, et al. Patient-reported outcomes in sickle cell disease and association with clinical and psychosocial factors: Report from the sickle cell disease implementation consortium. Am J Hematol. 2020;95(9):1066–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Hassell KL, Eckman JR, Lane PA. Acute multiorgan failure syndrome: a potentially catastrophic complication of severe sickle cell pain episodes. Am J Med. 1994;96(2):155–62. [DOI] [PubMed] [Google Scholar]
- 22.Fertrin KY, Costa FF. Genomic polymorphisms in sickle cell disease: implications for clinical diversity and treatment. Expert Rev Hematol. 2010;3(4):443–58. [DOI] [PubMed] [Google Scholar]
- 23.Steinberg MH. Predicting clinical severity in sickle cell anaemia. Br J Haematol. 2005;129(4):465–81. [DOI] [PubMed] [Google Scholar]
- 24.Piel FB, Jobanputra M, Gallagher M, Weber J, Laird SG, McGahan M. Co-morbidities and mortality in patients with sickle cell disease in England: A 10-year cohort analysis using hospital episodes statistics (HES) data. Blood Cells Mol Dis. 2021;89:102567. [DOI] [PubMed] [Google Scholar]
- 25.Hamideh D, Alvarez O. Sickle cell disease related mortality in the United States (1999–2009). Pediatr Blood Cancer. 2013;60(9):1482–6. [DOI] [PubMed] [Google Scholar]
- 26.DeBaun MR, Telfair J. Transition and Sickle Cell Disease. Pediatrics. 2012;130(5):926–35. [DOI] [PubMed] [Google Scholar]
- 27.de Montalembert M, Guitton C, Disease tFRCfSC. Transition from paediatric to adult care for patients with sickle cell disease. British Journal of Haematology. 2014;164(5):630–5. [DOI] [PubMed] [Google Scholar]
- 28.Rettenbacher E, Zaal J, Heijboer H, van der Plas EM, Hof M, Biemond BJ, et al. Mortality and Causes of Death From Sickle Cell Disease in The Netherlands, 1985–2017. J Pediatr Hematol Oncol. 2021;43(7):258–65. [DOI] [PubMed] [Google Scholar]
- 29.Mehari A, Alam S, Tian X, Cuttica MJ, Barnett CF, Miles G, et al. Hemodynamic predictors of mortality in adults with sickle cell disease. Am J Respir Crit Care Med. 2013;187(8):840–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Sachdev V, Kato GJ, Gibbs JS, Barst RJ, Machado RF, Nouraie M, et al. Echocardiographic markers of elevated pulmonary pressure and left ventricular diastolic dysfunction are associated with exercise intolerance in adults and adolescents with homozygous sickle cell anemia in the United States and United Kingdom. Circulation. 2011;124(13):1452–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Adam SS, Flahiff CM, Kamble S, Telen MJ, Reed SD, De Castro LM. Depression, quality of life, and medical resource utilization in sickle cell disease. Blood Adv. 2017;1(23):1983–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Levenson JL, McClish DK, Dahman BA, Bovbjerg VE, de ACV, Penberthy LT, et al. Depression and anxiety in adults with sickle cell disease: the PiSCES project. Psychosom Med. 2008;70(2):192–6. [DOI] [PubMed] [Google Scholar]
- 33.Villarroel MT E Symptoms of Depression Among Adults: United States, 2019. https://www.cdc.gov/nchs/products/databriefs/db379.htm: CDC; 2020. [PubMed] [Google Scholar]
- 34.Edwards G, Nuckols T, Herrera N, Danovitch I, Ishak WW. Improving Depression Management in Patients with Medical Illness Using Collaborative Care: Linking Treatment from the Inpatient to the Outpatient Setting. Innov Clin Neurosci. 2019;16(11–12):19–24. [PMC free article] [PubMed] [Google Scholar]
- 35.Alústiza JM, Artetxe J, Castiella A, Agirre C, Emparanza JI, Otazua P, et al. MR quantification of hepatic iron concentration. Radiology. 2004;230(2):479–84. [DOI] [PubMed] [Google Scholar]
- 36.Ballas SK. Iron overload is a determinant of morbidity and mortality in adult patients with sickle cell disease. Semin Hematol. 2001;38(1 Suppl 1):30–6. [DOI] [PubMed] [Google Scholar]
- 37.Fung EB, Harmatz P, Milet M, Ballas SK, De Castro L, Hagar W, et al. Morbidity and mortality in chronically transfused subjects with thalassemia and sickle cell disease: A report from the multi-center study of iron overload. Am J Hematol. 2007;82(4):255–65. [DOI] [PubMed] [Google Scholar]
- 38.Njoku F, Zhang X, Shah BN, Machado RF, Han J, Saraf SL, et al. Biomarkers of clinical severity in treated and untreated sickle cell disease: a comparison by genotypes of a single center cohort and African Americans in the NHANES study. Br J Haematol. 2021;194(4):767–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Farooq F, Mogayzel PJ, Lanzkron S, Haywood C, Strouse JJ. Comparison of US Federal and Foundation Funding of Research for Sickle Cell Disease and Cystic Fibrosis and Factors Associated With Research Productivity. JAMA Network Open. 2020;3(3):e201737–e. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from Department of Clinical Research, SSES, RTI International, Research Triangle Park, North Carolina, USA, upon reasonable request.