Treatment and prevention of pain due to vaso-occlusive crises in adults with sickle cell disease: an educational void

Sickle cell disease (SCD) affects approximately 72 000 individuals in the United States. Subjects with these disorders may suffer frequent vaso-occlusive crises (VOCs) marked by severe pain often requiring parenteral opioid administration in hospital emergency departments and inpatient units.¹  Guidelines for the management of pain due to VOCs in SCD, published in both the United States²  and Great Britain,³  have been available since 1999 and 2003, respectively. In the United States, these guidelines are also incorporated into a document on the management of SCD published by the National Institutes of Health (NIH) in 2002.⁴  However, patients and their families frequently express dissatisfaction with the care they receive in acute care settings.⁵  Moreover, while treatment with hydroxyurea was shown to decrease the incidence of VOCs in 1995, this medication is still underused.^6-8  These observations prompted a review of 19 textbooks in hematology, internal medicine, and emergency medicine published during or after the year 2003 to identify barriers to the effective management of VOCs in SCD.^9-27  (Specific chapters reviewed in each textbook are listed in Document S1, available on the Blood website; see the Supplemental Materials link at the top of the online article).

The 4 essential features of both guidelines are (1) rapid initiation of opioid therapy (within 15-30 minutes of arrival in the emergency department); (2) use of an adequate opioid starting dose; (3) frequent repeat doses of opioids (every 15-30 minutes) until pain is significantly improved; and (4) the need to select treatment regimens based on an individual's prior opioid-response history (Table 1).^2-4  The guidelines differ somewhat in their recommendations for the doses of opioid to be used when knowledge of prior opioid requirements is limited. Thus, the American Pain Society (APS) suggests an initial loading dose of morphine of 5 to 10 mg (or 1.5 mg hydromorphone) in adults weighing at least 50 kg with repeat doses being one quarter to one half of the loading dose given every 15 to 30 minutes with increasing dose titration to pain relief.^2,4  In contrast, the British Committee for Standards in Hematology (BCSH) suggests a loading dose of 0.1 mg/kg morphine with the same dose repeated every 20 minutes until effective analgesia is achieved.³  Both guidelines also indicate the need to monitor patients carefully for respiratory depression and other adverse effects of opioid therapy.

Unfortunately, most physicians are not familiar with these guidelines or with the treatment approaches they suggest. In an informal survey at our institution, none of 7 attending physicians on the adult hematology service (myself included) and none of 5 attending physicians in the emergency medicine department had any awareness of the existence of either guideline. Moreover, lack of awareness of guidelines may translate into inadequate patient care. Thus, in 2004, only 46% of 79 medical directors of American pain clinics indicated that opioids had a major role in the treatment of pain related to sickle cell disease, while 6% indicated that opioids were a last resort and 4% felt that opioids had no role whatsoever.²⁸  Similarly, in 2005, 25% of 109 physicians surveyed at 7 institutions with NIH-funded university-based comprehensive sickle cell centers indicated that a morphine dose of 2 mg intravenously every 4 to 6 hours was appropriate or excessive treatment for pain in a 66-kg man with SCD, while 33% thought that a morphine dose of 10 mg intravenously every 3 hours was either overtreatment or inappropriate.²⁹ 

The reason for this lack of knowledge is apparent from a review of the standard medical textbooks. As shown in Table 1, none of the reference works indicate a target time for the initiation of opioid therapy, and only 7 texts (37%) comment on the need to individualize therapy based on patient history. Suggested starting doses of morphine or hydromorphone are not given in 8 (42%) of the 19 references reviewed including 6 of the 9 hematology texts and 2 of the 3 emergency medicine texts. In the 11 other reference works, suggested starting opioid doses are consistent with or higher than guideline recommendations in 10 and lower than guideline recommendations in one.

The initial frequency of opioid administration is not suggested in 9 (47%) of these 19 references, only 3 of the 10 remaining texts suggest repeat opioid doses within a 20- to 30-minute interval, and 1 text suggests an interval of 30 to 45 minutes. Moreover, while one text notes that one half of the initial dose can be given 30 minutes after the first dose if pain is not adequately controlled, it is not clear whether additional doses can be given at 30-minute intervals for persistent poor pain control prior to resuming full-dose therapy every 2 hours.²⁴  Thus, overall, the frequency of opioid administration suggested is consistent with guidelines in 4 texts (21%) but only if this ambiguity is ignored and if the somewhat more extended interval of 30 to 45 minutes is accepted.²⁶ 

Only 7 texts cite either guideline in their lists of references.^{10,11,13-15,23,27}  Nonetheless, treatment recommendations are consistent with these guidelines in only 2 of them.^10,15 

VOC is often treated by physicians with little experience in acute pain management in other settings. Thus, the analgesic regimens suggested in the current guidelines may seem to require particularly high opioid doses administered at unusually frequent intervals. However, as shown in Table 2, while protocols used in different acute pain settings vary somewhat, maximum opioid administration allowed for in the first hour of treatment of patients with VOC is actually the same or less than that proscribed for the acute coronary syndromes, postoperative pain management, and the general treatment of acute pain in the emergency department.

Surprisingly little has been published about the actual opioid doses administered to adults with VOC. Gonzalez et al, using a protocol similar to the APS guideline, reported that 41.0 plus or minus 17.6 mg morphine was administered intravenously during emergency department visits by 12 adult sickle cell patients over a time frame of 5.5 plus or minus 1.6 hours.³¹  In our recent experience, there was a more than 10-fold interindividual variation in the total opioid requirement during the first 24 hours of admission of patients with VOC who were not on long-acting opioids (Table 3). Nonetheless, the mean opioid requirement in these patients was similar to that used postoperatively in representative studies after liver resection, caesarian section, total abdominal hysterectomy, knee surgery, and spinal surgery (Table 3).^32-36  This is particularly noteworthy since increased opioid metabolism has been demonstrated in the setting of SCD both in humans and in transgenic mice.^37,38  Although these comparisons are limited by the lack of uniform pain control end points and adjustment for patient weight, they provide a framework that suggests that opioid requirements in SCD are indeed proportional to those in other severe acute pain settings. Importantly, patients on chronic long-acting opioid therapy are likely to require higher opioid doses for the treatment of acute pain because of the development of tolerance and the occurrence of hyperalgesia.^4,39 

It is well recognized that objective measures of the presence and severity of pain are lacking. Moreover, studies of patients with advanced malignancy suggest that clinical caregivers frequently underestimate the severity of pain.⁴⁰  This has led to the core concept of pain treatment: “believe the patient!” Nonetheless, 86% of physicians in university hospital settings do not believe that self-report is the most reliable indicator of the existence and intensity of pain in patients with SCD.²⁹ 

Weighed against this concept is the pervasive and unfounded fear of drug addiction and drug-seeking behavior in acute and chronic pain patients in general and the SCD population in particular.^41-43  For example, 26% of hematologists and 53% of ED physicians in one survey believed that at least 20% of SCD patients are addicted.⁴²  Moreover, health care personnel consistently indicated that SCD patients were almost twice as likely to be opioid dependent as other pain patients presenting to an emergency department.⁴¹  In fact, however, estimates of the incidence of this problem in SCD patients parallel those in the community at-large (Table 4).^44-48  Thus, 8 (8.3%) of 96 patients older than 18 years in the adult SCD program at Yale New Haven Hospital have exhibited behaviors consistent with substance abuse. This incidence is similar to that reported in the population as a whole in New Haven in 1984 and in Connecticut in 2004 to 2005.^47,48  Moreover, many behaviors construed as indicative of addiction in SCD patients often result from undertreatment of pain (ie, pseudoaddiction).⁵⁰  Finally, racial prejudices also contribute to ineffective painmanagement in this population.⁵¹  These misconceptions and prejudices can lead to clinician behaviors that seriously compromise patient care.⁵² 

It is unfortunate then that only 7 (37%) of the 19 texts reviewed provide reassurance of a low risk of opioid addiction in sickle cell patients (Table 5). In contrast, sections on pain treatment in advanced cancer were present in 12 texts, and reassurance of the low risk of opioid addiction in this setting is provided in 11 of them (91%) (χ² = 8.941; P < .005). Similar reassurance is also provided in the chapters on acute pain management in all 3 emergency medicine texts.^18-20 

The Multicenter Study of Hydroxyurea (MSH) published in 1995 treated patients with at least 3 severe VOCs/year with hydroxyurea at an initial dose of 15 mg/kg per day.⁶  This dose was increased by 5 mg/kg per day every 12 weeks up to a maximum of 35 mg/kg per day as long as the neutrophil count exceeded 2 × 10⁹/L. As shown in Table 6, only 8 (50%) of the 16 hematology and internal medicine texts provided a reasonable indication for hydroxyurea therapy and this was vague in 6 of them; only 8 texts (50%) gave an initial dose of hydroxyurea consistent with the published studies; only 4 texts (25%) provided an appropriate schedule for increasing the dose of hydroxyurea; and only 5 texts (31%) noted the need to maintain the neutrophil count higher than 2 × 10⁹/L. Moreover, only 2 texts (13%) met all 4 of the criteria listed. (Emergency medicine texts were excluded from this analysis.)

Treatment recommendations for von Willebrand disease (VWD), hemophilia A, β-thalassemia major (β-Thal), thrombotic thromobocytopenic purpura (TTP), idiopathic thrombocytopenic purpura (ITP), and acute myelogenous leukemia (AML) were also reviewed. Textbook recommendations were considered consistent with current standards of practice if they met all of the recommendations for each disorder (Table S1). As shown in Table 7, with the exception of VWD, these disorders have incidence rates or prevalence rates significantly lower than that of SCD. Nonetheless, treatment recommendations consistent with generally accepted standards of care are given in 53% to 84% of texts. Overall, textbooks were significantly less likely to meet guidelines for both opioid and hydroxyurea therapy in SCD than treatment standards in all the other hematologic disorders reviewed (χ² > 4.07; P < .05).

Since many (albeit not all) physicians involved in the care of adults with SCD have access to online electronic medical journal publications (usually in university settings), these resources were also reviewed. Searches were performed using “the disease name + therapy/therapeutics” as keyword search terms with Medline/PubMed, which is universally available on the National Medical Library website (http://www.nlm.nih.gov).⁵³  This consistently resulted in more than 1200 “hits” for each disorder. Therefore, to make these searches more usable, references were then limited to review articles involving human subjects that were written in English and that appeared in core medical journals during the last 10 years. Pediatric journals were excluded except for the treatment of thalassemia and hemophilia A. The findings paralleled those in the medical textbooks. Thus, as shown in Table 8, none of the 4 review articles identified provided adequate information on opioid use in VOC and only 1 of 9 review articles provided sufficient information on hydroxyurea therapy. Moreover, the BCSH guideline paper was not identified by this search strategy.³  In contrast, reviews meeting the treatment criteria in Table S1 were easily identified for all other hematologic disorders evaluated with the exception of hemophilia.

Although randomized controlled trials have not been conducted, use of guideline recommendations appears to improve the quality of care of patients with VOC by shortening the time to pain relief and decreasing the need for hospitalization when used in both specialized day hospital units and emergency room settings.^54-56  In fact, even more intensive regimens may be warranted. Thus, a dose of 0.1 mg/kg morphine intravenously has recently been shown to have only limited effectiveness at 30 minutes in treating acute pain in the emergency room.⁵⁷ 

Although not easily accessible, comprehensive discussions of pain in SCD and its management have been written by both Ballas⁵⁸  and Benjamin.⁵⁹  Much of the information in the former monograph can be found in guideline form through the City of Hope Pain and Palliative Care website (http://www.cityofhope.org/prc/). The 2002 NIH monograph on the management of sickle cell disease⁴  can also be obtained at both the City of Hope website and the Sickle Cell Center Information website of the Georgia Comprehensive Sickle Cell Center at Grady Health System (http://www.scinfo.org).

Patients with SCD residing in rural areas underutilize services in urban sickle cell centers.⁶⁰  Thus, it is important for current guideline recommendations to be readily available to all physicians in family practice, internal medicine, hematology, and emergency room settings. Nonetheless, most standard medical texts provide neither adequate information for the treatment or prevention of pain due to vaso-occlusive crisis in SCD nor reassurance of the unlikelihood of addiction in this population. Nor is this need met by easily obtained review articles in core medical journals. Modern textbooks face a daunting task of incorporating a rapidly expanding body of knowledge into a usable resource that defines the pathogenetic, clinical, and therapeutic aspects of a great many disorders. It is important therefore to review the success with which this goal is met and to direct readers to other readily available resources when space and scope limitations exist. Consideration should also be given to where essential primary source material can be found. While the BCHS recommendations appear in a major hematology journal, this article is not identified by a Medline search for relevant review articles. Moreover, the APS/NIH guidelines exist only as stand-alone publications. Education directed at both increasing awareness of pain treatment guidelines and understanding the behaviors and prejudices of health care professionals that compromise patient care in this setting is greatly needed. Finally, while available guidelines provide frameworks for treatment of VOC in SCD, further research is required to define optimum methods for analgesic intervention in this disorder.⁶¹ 

I thank E. Doram, LCSW, and G. Nelson, APRN, for their helpful comments and insights.

Contribution: L.R.S. is solely responsible for the collection and analysis of the data and for the preparation of this paper.

Conflict-of-interest disclosure: The author declares no competing financial interests.

Correspondence: Lawrence R. Solomon, Hematology Section, Department of Medicine, Yale University School of Medicine, 403 www; 333 Cedar St, PO Box 208201, New Haven, CT 06520-8021; e-mail: lawrence.solomon@yale.edu.