An impact evaluation of two modes of care for sickle cell disease crises

Approximately 100,000 Americans currently live with a group of inherited red blood cell disorders known as sickle cell disease (SCD). Individuals with SCD commonly experience prolonged and painful vaso-occlusive crises (VOCs) that require hospitalization. These painful crises, the most frequent complication of SCD, occur sporadically and last many days. Patients often seek treatment in emergency departments (EDs), where they receive intravenous infusions of pain-reducing medicines. However, VOC patients have expressed dissatisfaction with the care they receive in EDs, citing significant delays in the administration of required medication [1]. For instance, Tanabe et al. found a median wait time of 90 min for the first dose of which is analgesic, significantly longer than the 15–20 min that the American Pain Society recommends [1]. Significant differences in time to first dose across patients reporting similar pain levels were also found. Silbergleit et al. found just 20% of ED physicians follow a protocol when caring for patients experiencing a VOC [2]. These deficiencies have encouraged the creation of specialty centers for handling VOCs.

We developed a population-level Markov model to simulate the effects of uptake of IC care for VOCs over a 10-year time horizon. Our model estimates the impact of a baseline and ‘'treated' state. In the baseline state, all SCD patients who experience VOC events and require acute care, receive their care in an ED. In the treated state, patients receive acute care in either the ED or an IC. The number of acute care events in the IC in the treated state is based on assumptions of the proportion of patients with access to ICs and the percentage of patients seeking care at an IC when an IC is available.

Each year, a new cohort of SCD patients enters the model. In the treated state, these patients receive care in the ED or IC, or conduct self-care at home, or receive no care. To simulate the patient population with access to an IC, we estimate a beginning estimate of patients with access, followed by an annual uptake rate of IC access every year. A recent literature has outlined the elements needed to run an adult specialty sickle cell center and highlighted implementation workshops with new SCD care centers [7]. The Patient-Centered Outcomes Research Institute (PCORI) is funding work to identify barriers to the expansion of specialty care for SCD and develop an implementation toolkit [8]. To account for this push to expand access to adult sickle cell disease specialty care, we assume a constant rate of growth in patient access to ICs over time. The model estimates acute care event frequency, payer and out-of-pocket medical costs, caregiver time and travel costs and employer costs in the form of foregone wages.

The 10-year healthcare utilization and cost impacts of having access to specialty ICs for the treatment of SCD/VOC pain crises compared with the current treatment practice at EDs are shown in Table 1. We estimated a 14% reduction in ED visits and a 12% reduction in hospital admissions. Based on our model, the patients who avoided 175,000 ED visits and 55,000 hospital stays instead had nearly 205,000 treatments at specialty ICs. Total societal cost savings are the sum of formal and informal sector cost savings, equaling US$2,007,153,548 over 10 years, or US$3,275 per patient per year.

Because of the high cost of the avoided ED visits and hospital stays, payers would receive the vast share of the cost savings estimated from increased access to ICs for treatment of SCD crises. On the 10-year time horizon, payers save US$1,811,386,274. This equates to a per-patient savings of US$2956 per year.

Patients benefit from treatment at ICs compared with EDs regarding reduced time and out-of-pocket costs. Based on our model assumptions of less patient time for treatment at ICs, we estimate that over 10 years, patients would save 6,385,842 h compared with the treatment of all patients at EDs (10 h per patient per year). Patients would also have lower out-of-pocket healthcare costs when they have access to ICs. On the 10-year time horizon, patients would save US$170,014,834 in out-of-pocket medical costs, or US$277 per patient per year.

The reduced time for treatment at ICs also translates to benefits for employers. We estimate 870,603 fewer h of missed work. From the employer's perspective, this equates to a US$25,752,439 in avoided productivity losses.

Patients who receive treatment for SCD pain crises often require accompaniment at completion of their treatment. We model the time and time costs from the family perspective by assuming patients are accompanied by one working-age care partner while traveling to and from treatments of SCD pain crises. Because many ICs are co-located with hospitals, we assume equivalent travel distance and times to both EDs and ICs. As a result, we estimate no difference in family time or family costs between baseline treatment state and the alternative, IC intervention state.

We estimated that there are no transportation savings or cost differences between the baseline and treatment states. In our model, all patients who seek care make two trips: one to the acute care facility and another to their home after discharge. Patients experience these two trips regardless of whether they receive their initial care in the IC or ED. Our assumption is that IC care is embedded in a larger clinical space shared by other programs, such as a hospital. This assumption mirrors that of many existing models of SCD care [7].

Our analysis follows growing findings of the efficacy of specialty SCD care versus care received in the ED. This study is, to the authors' knowledge, the first to compare the economic impacts of specialty SCD care with traditional care at the societal level. Researchers have estimated the annual economic burden of SCD to payers and patients at US$2.98 billion per year [24]. Lee and colleagues [15] conducted a systematic review of the literature on the economic burden of adults with SCD in the USA and found that the most current literature is related to healthcare resource utilization, and only some of the recent literature identifies direct medical costs, highlighting the need for more economic research in this area. Molokie et al. [25] compared inpatient admissions and lengths of stay following care in specialty care centers and the ED. This research found a higher rate of hospital admission following ED care and similar lengths of stay in the hospital following ED and specialty care.

Other benefits of expanded access to specialty ICs are not measured in our impact estimates. These include reported patient satisfaction, decreased time to first dose, and more frequent pain reassessment throughout the patient care experience [3]. This suggests that specialty care could reduce costs and utilization while increasing patient satisfaction.

One recent study examined the relationship between prescription drug monitoring programs (PDMPs) and access to opioids for patients in the ED. Zhang et al. [26] found a reduction in opioid dispensation for patients presenting to the ED with a SCD crisis after the implementation of PDMPs [26]. These PDMPs could potentially affect our model inputs and results by increasing the time to first dose of pain medication in EDs, resulting in lower quality of care. Additionally, PDMPs could have a greater impact on patients presenting to the ED because of less provider familiarity with SCD crises and the individual patient.

It is worth noting that we do not model an immediate nationwide or universal rollout of specialty care. Our model assumes a gradual uptake of specialty care over 10 years. If a steady state of widespread specialty care use was achieved more quickly than we assumed, then our estimated savings would increase.

The United States could potentially avert 55,086 inpatient hospital stays with increased uptake of specialty care for SCD crises over 10 years. By providing specialty care and reducing hospital admissions, payers and patients could save over US$1.9 billion in formal medical costs over 10 years. This equates to approximately 6.4 million h of patient time, or US$26 million in productivity losses as a result of reduced employee absenteeism. These savings equate to approximately US$2.0 billion in total societal costs.

We gratefully acknowledge funding for this study from the Patient-Centered Outcomes Research Institute. We thank J Ogunsile, S Lanzkron and T Hoerger for their valuable feedback.

The contents of this work are solely the responsibility of the authors and do not necessarily represent the views of the Patient- Centered Outcomes Research Institute (PCORI), its Board of Governors or Methodology Committee.

This work was funded by the Patient-Centered Outcomes Research Institute (PCORI) under contract IDIQ-SOW #16-INSIGHT-AOSEPP-ENG. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Writing assistance in terms of editing and formatting of this manuscript was provided in-house within the terms of the contract.