Abstract
Purpose: To assess current physician self-reported practices regarding initial management of childhood idiopathic thrombocytopenic purpura (ITP) and to determine physician self-reported willingness to participate in randomized clinical trials comparing different initial management strategies.
Patients and Methods: A questionnaire was mailed in November 1997 to all 720 members of the American Society of Pediatric Hematology/Oncology asking how they would diagnose and manage ITP in children 18 months, 5 years, and 15 years of age who were experiencing either dry purpura (cutaneous hemorrhage only) or wet purpura (active mucous membrane hemorrhage). Specific questions dealt with bone marrow aspiration, hospital admittance, treatment strategy, and specific doses of corticosteroids and intravenous immunoglobulin.
Results: The response rate to the questionnaire was 57%. Most respondents indicated they usually perform a bone marrow aspirate when corticosteroids are to be prescribed and administer drug therapy to patients with newly diagnosed ITP with wet or dry purpura. Only 16% of respondents would administer no drug therapy to a child with dry purpura. Intravenous immunoglobulin (IVIG) was preferred to steroids, with anti-D immunoglobulin prescribed less frequently. Hospital admittance often was used for patients with dry purpura and usually recommended for patients with wet purpura. Most respondents expressed willingness to randomize patients with dry purpura to IVIG versus no therapy and those with wet purpura to IVIG versus prednisone as part of a randomized controlled clinical trial.
Conclusions: The self-reported care of the patient with ITP was influenced by the severity of presentation (dry versus wet purpura). Most physicians reported they would administer specific drug treatment in both scenarios. This survey illustrates the diverse diagnostic and management strategies currently used in childhood ITP. Because no one therapeutic approach is predominant and a scientific basis for decision making in childhood ITP has not been developed, future randomized trials are warranted. On the basis of these survey results, such trials are desired by most pediatric hematology/oncology specialists.
The optimal strategy for diagnosing and treating idiopathic thrombocytopenic purpura (ITP) during childhood is a matter of controversy (1-3). Current clinical approaches remain based on personal experience, uncontrolled case series, and several randomized clinical trials with short follow-up and in which platelet count was the only outcome measure. Two expert panels have developed ITP practice guidelines in the United Kingdom and the United States (4,5), but the recommendations were based on opinion, not data, and the degree to which the recommendations are followed by practicing pediatric hematologists is unclear. A recent report of current practice patterns from the United Kingdom indicated little compliance with published guidelines (6).
Patterns of practice can be assessed by a questionnaire/survey, asking practitioners to respond to questions about hypothetical situations or cases. Such a survey of pediatric hematologists regarding ITP practices was conducted in 1986 (7), but this was before intravenous immunoglobulin (IVIG) and anti-D immunoglobulin were used. Thus, the current authors elected to conduct a survey regarding contemporary diagnostic and treatment strategies. The target audience was the membership of the American Society of Pediatric Hematology/Oncology, which includes most pediatric hematology/oncology specialists in the United States and practitioners from 40 other countries. The purpose of this communication is to describe the results of the survey.
MATERIALS AND METHODS
A survey describing diagnostic and treatment approaches regarding acute childhood ITP was developed and sent by regular mail to all 720 members of the American Society of Pediatric Hematology/Oncology (ASPH/O) in November 1997. Most of the 1,100 to 1,200 pediatric hematology/oncology specialists estimated to be practicing in the United States are members of the organization. The covering letter, signed by the Society's president and vice president (two of the authors, A.C. and G.B.) informed respondents that approximately 15 minutes would be required to complete the six-page survey, and a self-addressed envelope was enclosed. Approximately 2 weeks later, a postcard was mailed to each ASPH/O member, reminding them of the response deadline 2 weeks later.
The survey instrument consisted of two clinical case scenarios of a typical patient with acute ITP with a platelet count of 3,000 per mm3. One case described a child with only cutaneous hemorrhage (dry purpura), and the other described a patient with ongoing mucous membrane bleeding and anemia (wet purpura). The scenarios are summarized in Table 1. For both dry and wet purpura scenarios patients 18 months, 5 years, and 15 years of age were chosen. Except for age and type of bleeding, the dry and wet purpura cases were identical. Thus, six scenarios were presented. The age of 5 years was considered to represent the most typical age of an affected child with ITP. The 18-month age was chosen to reflect that toddlers often stumble and hit their heads and thus might be viewed as being at greater risk of intracranial hemorrhage. This perceived risk might alter a practitioner's approach from that used for a 5-year-old patient. A 15-year-old patient was chosen in view of the different activity level and possibly greater risk of serious hemorrhage in teenagers. In addition, the similarity of ITP in teen-agers to that in affected adults might be expected to affect care decisions.
[Help with image viewing]
[Email Jumpstart To Image]
|
TABLE 1. Idiopathic thrombocytopenic purpura clinical scenarios and two types of presentation
|
For four of the six case scenarios (dry purpura at 18 months, 5 years, and 15 years of age, and wet purpura at 5 years of age), questions were asked regarding whether the respondent would admit to the hospital such a patient, perform a marrow aspirate regardless of therapy, or perform a marrow aspirate if corticosteroids were to be given. The respondent also was asked whether he or she normally would treat such a patient with a pharmacologic agent, and if so, what type of treatment would be administered. For corticosteroids and IVIG, respondents were asked which of several possible dosage regimens and schedules would be used. In the wet purpura cases, the respondent was asked whether a platelet transfusion also would be given. None of the questions were open ended; a specific response (one of several choices) was requested in each case.
Respondents were queried about the number of years they have been practicing hematology/oncology, the number of patients with ITP each year that their institution and they personally see, and whether or not they practice primarily hematology, oncology, or both subspecialties. Respondents also were asked whether they would be interested in participating in controlled clinical trials involving ITP, including randomization of patients with dry and wet purpura to alternative treatment regimens.
The survey results were entered into Microsoft Access (Redmond, WA, U.S.A.) software and analyzed with SAS (Cary, NC, U.S.A.) software. The X2 test, Fisher exact test, and Bowker test of symmetry were used to assess significance.
RESULTS
Number and Demographics of Respondents
There were 409 responses to the questionnaire/survey (57% response rate). The response rate by census division in the United States ranged from 46% to 62%. The response rate outside the United States was 54%. Thirty-nine respondents did not complete the questionnaire because they indicated having no direct involvement with patients with ITP. The remaining 370 questionnaires are included in the subsequent analyses. Characteristics of the respondents whose questionnaires were analyzed are depicted in Table 2. There was an equal distribution between those practicing pediatric hematology/oncology less than 10 years, from 11 to 20 years, and greater than 20 years. As expected, most respondents had full-time academic positions and practiced a mix of hematology and oncology. One hundred eleven (33%) of the respondents reported they saw 10 or more patients with ITP per year. Physicians who reported seeing 10 or more patients with ITP per year were more likely to report their practice was primarily hematology (P < 0.05).
Bone Marrow Aspiration
As shown on Table 3, most respondents indicated they usually or always perform a bone marrow aspirate when corticosteroids are to be prescribed to the hypothetical patient with ITP (i.e., regardless of the age of the patient or nature of the purpura, approximately 75% of the respondents would always or usually perform a marrow aspirate). However, when corticosteroids are not going to be prescribed, a marrow aspirate is performed much less frequently. The influence of corticosteroid therapy (to be given or not to be given) on marrow aspirate practice was statistically significant (P <= 0.001).
[Help with image viewing]
[Email Jumpstart To Image]
|
TABLE 3. Self-reported performance of bone marrow aspiration for diagnosis of idiopathic thrombocytopenic purpura according to bleeding symptoms and age of patient
|
A 5-year-old boy with wet purpura is more likely to have a bone marrow examination, in the absence of steroid therapy, than is a 5-year-old patient with dry purpura. As shown in Table 3, 21% of respondents not prescribing corticosteroids always or usually perform a marrow aspirate on the 5-year-old with wet purpura, in contrast to just 9% when the purpura is dry (P < 0.001). Physicians who have been in practice the longest are more likely to perform a marrow aspirate in the absence of corticosteroid therapy. For example, for the 5-year-old patient with dry purpura, only 3% of physicians in practice 10 years or less always or usually perform a marrow aspirate, whereas 14% of practitioners with more than 20 years experience perform marrow aspirates under such conditions (P = 0.004).
Hospital Admittance
Table 4 depicts self-reported hospital admittance practices. Whereas 85% of respondents would always admit a 5-year-old patient with wet purpura, only 41% generally admit patients with dry purpura (P <= 0.001). When dry purpura cases are considered, teenagers are less likely (P <= 0.001) and toddlers more likely (P <= 0.001) to be admitted to the hospital is than the 5-year-old.
[Help with image viewing]
[Email Jumpstart To Image]
|
TABLE 4. Self-reported practice of hospitalization upon initial diagnosis of idiopathic thrombocytopenic purpura according to bleeding symptoms and age: percentage of respondents recommending hospitalization
|
Treatment
Treatment decision making is summarized in Table 5. Most respondents indicated they would administer drug therapy to patients with newly diagnosed ITP, irrespective of the patient's age or type of bleeding. Regarding the 5-year-old patient with dry purpura, only 16% of respondents chose no drug therapy, 19% corticosteroids, 45% IVIG, 10% anti-D, and 10% other options. Virtually all of the respondents recommended drug therapy for the hypothetical 5-year-old patient with wet purpura. Here again, IVIG was the preferred therapy, chosen by 60% of the respondents. For the 18-month-old and 15-year-old patients with dry purpura, the most common self-reported treatment was IVIG, followed by corticosteroids and anti-D. Among patients with dry purpura, IVIG is prescribed less often and no drug therapy more often for the 5-year-old patient compared with the 18-month-old patient (P = 0.0001).
[Help with image viewing]
[Email Jumpstart To Image]
|
TABLE 5. Self-reported initial treatment of children with idiopathic thrombocytopenic purpura according to the patient's bleeding symptoms and age of patient
|
The most frequently self-reported IVIG dose was 1 g/kg, repeated the next day if the platelet count was less than 30,000 per mm3 (Table 6). Only 9% to 11% of respondents chose a lower dose of 0.8 g/kg or a single dose of 1.0 g/kg. For patients with dry purpura, there were no differences among the self-reported IVIG doses used for patients of various ages. A second dose of 1.0 g/kg, irrespective of the next day's platelet count, was more likely to be ordered if the 5-year-old patient had wet purpura (P = 0.05).
[Help with image viewing]
[Email Jumpstart To Image]
|
TABLE 6. Self-reported intravenous immunoglobulin dose for children with idiopathic thrombocytopenic purpura according to type of purpura and age of patient
|
As shown in Table 7, the most frequently prescribed steroid dose was prednisone, 2 mg/kg per day orally for 7 or more days. The 5-year-old patient with wet purpura was more likely to receive a larger oral dose or a parenteral corticosteroid regimen (P = 0.002). There was no apparent effect of age on the corticosteroid regimen used for patients with dry purpura.
[Help with image viewing]
[Email Jumpstart To Image]
|
TABLE 7. Self-reported steroid dose for children with idiopathic thrombocytopenic purpura according to type of purpura and age of patient
|
Assessment of physician experience significantly influenced some management decisions. Of physicians who have practiced less than 10 years or 10 to 20 years, 12% or 11%, respectively, would administer no drug therapy. However, of practitioners with more than 20 years of experience, 24% would not treat the 5-year-old patient with dry purpura with a pharmacologic agent (P = 0.009). When comparing physicians who reported they saw 10 or more patients with ITP per year with physicians who reported they saw fewer than 10 such patients a year, there was no difference in the proportion of physicians who chose some treatment versus no treatment, but there was a significant difference for preferred initial treatment choice. For the 5-year-old patient with dry purpura, 16% of respondents seeing fewer than 10 patients with ITP per year self-reported using steroids, 50% IVIG, and 8% Anti-D. Among those seeing 10 or more patients with ITP each year, 27% chose steroids, 34% IVIG, and 15% Anti-D (P = 0.01). The same trend of initial treatment choices also was significant (P < 0.05) for the other three patient scenarios. In addition, for each of the four patient scenarios, the physicians who reported seeing 10 or more patients with ITP per year chose higher steroid doses for initial treatment (prednisone, 4 mg/kg/day or methylprednisolone 30 mg/kg/day v prednisone 2 mg/kg/day) than did physicians who saw fewer patients with ITP (all P < 0.05).
Platelet Transfusion
For the 5-year-old patient with wet purpura, respondents were asked whether or not they would administer a platelet transfusion in addition to or rather than other therapy. Ninety-four percent (343 of 366) of respondents reported they would not administer a platelet transfusion under such circumstances.
Attitudes About Participation in Randomized Trials of Alternative Initial Treatments
Respondents were asked whether or not they would be willing to randomize a 5-year-old patient with dry purpura to no drug therapy (i.e., observation) versus IVIG, 1 g/kg, repeated the next day if the platelet count were 30,000 per mm3, as part of a National Institutes of Health funded multicenter clinical trial. Sixty-eight percent of respondents expressed willingness to randomize their patient under such circumstances. This response is in contrast with only 16% of the respondents who would choose no drug treatment as their management choice outside of the clinical trial setting (Table 5). Respondents also were queried regarding their willingness to randomize a 5-year-old patient with wet purpura to prednisone 4 mg/kg per day for 7 days versus IVIG 1 g/kg, repeated the next day if the platelet count were less than 30,000 per mm3. Here again most (79%) of the respondents indicated their interest in participating in such a study, although only 11% would, in the absence of a clinical trial, use prednisone alone (without IVIG) in such patients (Table 5).
DISCUSSION
Several randomized controlled studies of ITP treatments conducted during the past few years have shown that IVIG raises the platelet count more rapidly than when no treatment is given and suggest that IVIG may be more effective than prednisone in promptly increasing the platelet count (8,9). However, with higher doses of prednisone (4 mg/kg/day) or with anti-D therapy, the platelets increase nearly as rapidly as with IVIG therapy (10). Even without drug therapy, the platelet count usually increases within days to several weeks (9,11). It is widely assumed that a prompt increase in the platelet count is desirable in patients with ITP to prevent serious hemorrhage. However, no therapeutic modalities have been shown to prevent intracranial or other serious bleeding (1,2). A recent study performed by Medeiros and Buchanan (12) focused on patients with ITP who were experiencing major hemorrhage. The study results suggested only a small minority of such children with more serious bleeding had a prompt (by the next day) increase in platelet count after IVIG or corticosteroid therapy (12). Thus, it is not known if therapeutic intervention would prevent major hemorrhage, so additional randomized trials, with clinical endpoints (such as type and severity of bleeding, cost assessment, and quality of life measures, in addition to platelet count) need to be conducted.
Until future clinical trials provide a sound basis for ITP management, one must rely on recommendations based on opinion. A recent report by Bolton-Maggs and Moon from the United Kingdom examined the diagnostic evaluation and therapy of all children presenting with ITP during a 14-month period (4,6) and contrasted their findings with the opinion-based diagnostic and management guidelines published 5 years previously by an expert panel of British pediatric hematologists (4). The experts recommended that pharmacological therapy was not required in the absence of serious bleeding, regardless of the platelet count; that prednisone was preferred to IVIG if therapy was used; that routine hospital admittance was unnecessary; and that a bone marrow aspirate should be performed before corticosteroid therapy (4). However, actual practice results were much different (6). Eighty percent of the 435 children with ITP included in the survey were admitted to the hospital (although only 10% had serious bleeding), 75% were treated, IVIG was administered more often than was prednisone, larger-than-recommended doses of prednisone were routinely used, and 30% of children receiving corticosteroids failed to have a prior marrow aspirate.
In the current survey, the authors aimed to determine what factors (patient age and extent of bleeding) affected self-reported clinical decision making for the diagnosis and management of ITP in children presenting with a platelet count of 3,000 per mm3. The last published survey of ITP management practices in the United States reflected viewpoints from the mid-1980s (7), before IVIG was in widespread use and before newer corticosteroid regimens (using methylprednisolone or dexamethasone) and anti-D treatment were developed. In that earlier study, performance of a marrow aspirate to rule out leukemia and prescription of corticosteroids were commonly used (7).
One interesting finding was the self-reported choice of corticosteroid and IVIG dose used by respondents who chose treatment. Prednisone at 2 mg/kg/day and IVIG at 1 g/kg (repeated the next day if the platelet count remained below 30,000 per mm3) were the regimens chosen most often (Tables 6 and 7), despite recent randomized trials demonstrating greater or equivalent efficacy of 0.8 g/kg of IVIG as a single dose or 4 mg/kg/day of prednisone in increasing the platelet count (9,10).
Not surprisingly, the hypothetical patients with ongoing mucous membrane bleeding were treated, by physician self-reporting, more aggressively than were the patients with dry purpura; the former virtually always received some type of treatment, were more likely to receive IVIG therapy or combined corticosteroids and IVIG, and in 6% of cases were viewed as candidates for a platelet transfusion. Overall, the age of the patient had little impact on self-reported treatment decisions. However, the 18-month-old patient was somewhat more likely to be admitted to the hospital and receive drug therapy, probably because of a perceived greater risk of hemorrhage in view of the inability to control the toddler's activity level.
Most respondents did not think it necessary to perform a bone marrow aspirate as long as corticosteroids were not to be given. This result might reflect several recent studies showing that leukemia is highly unlikely to present as isolated thrombocytopenia (13,14). There were few significant relationships between self-reported management decisions and the respondents' type of practice (hematology, oncology, or combined) or number of years in practice. The authors collected this information because we speculated that more experienced physicians and those who focus primarily on hematology might be less aggressive in their management decisions-less likely to routinely perform a marrow aspirate and i.e., less likely to treat patients with dry purpura.
Although only 16% of the respondents reported that no drug treatment was their choice of care for a patient with dry purpura, 68% of the physician respondents reported willingness to randomize such a patient to either no drug treatment or IVIG. In the hypothetical 5-year-old patient with wet purpura, only 11% reported that prednisone alone was their treatment choice. However, 79% of physicians were willing to randomize such a hypothetical 5-year-old patient with wet purpura to either prednisone or IVIG. Thus, it is clear that many ASPH/O members are interested in participating in randomized studies of the initial management of childhood ITP.
This study has several limitations. As with any survey of physician self-reporting practices, the behavior when encountering an actual patient may differ from that described in response to a questionnaire/survey regarding a hypothetical patient. Second, the respondents might not be the physicians actually diagnosing and caring for children with ITP. However, the ASPH/O membership is representative of pediatric hematology/oncology specialists in the United States, where most patients with ITP are referred to subspecialists for consultation or care. Third, one might question the current relevance of one key result (the frequent use of IVIG for children with ITP) whether they have dry or wet purpura. This survey was undertaken late in 1997, just as a major national shortage of IVIG was developing. Since then, the IVIG shortage has become even more pressing, and one recent report from the Texas Children's Hospital describes a marked reduction in IVIG use in patients with ITP as a result (15). The questionnaire did not ask what management decision would be used if IVIG were not available. Fourth, only 57% of surveyed physicians responded. However, this is consistent with the general rule that 40% of a survey population responds to each mailing (16). Thus, the survey, with one mailed reminder, would have at most an anticipated 64% response. Finally, it is possible the platelet count of 3,000 per mm3 used in the case scenarios was a predominant factor in therapeutic decision making. For instance, had a platelet count of 8,000 per mm3 been chosen, would a larger proportion of the respondents have chosen not to administer drug treatment to the child with dry purpura?
In conclusion, the results of this questionnaire/survey suggest that most pediatric hematology/oncology specialists in the United States would manage ITP in a child with dry purpura by not performing a bone marrow examination and administering drug treatment, preferably IVIG, to raise the platelet count. The patient with wet purpura would receive similar care, but would more frequently receive IVIG or combined corticosteroid and IVIG therapy. A substantial majority of respondents stated their willingness to randomize patients with either dry or wet purpura to alternative treatments if such care were undertaken as part of a controlled clinical trial. The significant challenge is to develop sound studies with clinically important outcome measures that will allow for treatment of ITP to be evidence based, rather than determined by studies focusing only on short-term platelet count responses and on personal experience.
REFERENCES
1. Buchanan GR. Childhood acute idiopathic thrombocytopenic purpura: How many tests and how much treatment required? J Pediatr 1985;106:928-30. Bibliographic Links [Context Link]
2. Lilleyman JS. Definitions and dogma in childhood idiopathic thrombocytopenic purpura. Pediatr Hematol Oncol 1993;10:11-4. [Context Link]
3. Modak SI, Bussel JB. Treatment of children with immune thrombocytopenic purpura: Are we closer to resolving the dilemma? J Pediatr 1998;133:313-4. Ovid Full Text Full Text Bibliographic Links [Context Link]
4. Eden OB, Lilleyman JS on behalf of the British Paediatric Haematology Group. Guidelines for management of idiopathic thrombocytopenic purpura. Arch Dis Child 1992;67:1056-8. Bibliographic Links [Context Link]
5. George JN, Woolf SH, Raskob GE, et al. Idiopathic thrombocytopenic purpura: A practice guideline developed by explicit methods for the American Society of Hematology. Blood 1996;88:3-40. Bibliographic Links [Context Link]
6. Bolton-Maggs PHB, Moon I. Assessment of UK practice for management of acute childhood idiopathic thrombocytopenic purpura against published guidelines. Lancet 1997;350:620-3. Full Text Bibliographic Links [Context Link]
7. Dubansky AS, Oski FA. Controversies in the management of acute idiopathic thrombocytopenic purpura: A survey of specialists. Pediatrics 1986;77:49-52. Full Text Bibliographic Links [Context Link]
8. Imbach P, Berchtold W, Hirt A, et al. Intravenous immunoglobulin versus oral corticosteroids in acute immune thrombocytopenic purpura in childhood. Lancet 1985;2:464-8. Bibliographic Links [Context Link]
9. Blanchette VS, Luke B, Andrew M, et al. A prospective, randomized trial of high-dose intravenous immune globulin G therapy, oral prednisone therapy, and no therapy in childhood acute immune thrombocytopenic purpura. J Pediatr 1993;123:989-95. Bibliographic Links [Context Link]
10. Blanchette V, Imbach P, Andrew M, et al. Randomized trial of intravenous immunoglobulin G, intravenous anti-D, and oral prednisone in childhood acute immune thrombocytopenic purpura. Lancet 1994;344:703-7. Full Text Bibliographic Links [Context Link]
11. Buchanan GR, Holtkamp CA. Prednisone therapy for children with newly diagnosed idiopathic thrombocytopenic purpura. Am J Pediatr Hematol Oncol 1984;6:355-61. Ovid Full Text Bibliographic Links [Context Link]
12. Medeiros D, Buchanan GR. Major hemorrhage in children with idiopathic thrombocytopenic purpura: Immediate response to therapy and long-term outcome. J Pediatr 1998;133:334-9. Ovid Full Text Full Text Bibliographic Links [Context Link]
13. Dubansky AS, Boyette J, Falletta J, et al. The rarity of isolated thrombocytopenia in acute lymphoblastic leukemia. Pediatrics 1989;84:1068-71. Full Text Bibliographic Links [Context Link]
14. Calpin C, Dick P, Poon A, et al. Is bone marrow aspiration needed in acute childhood idiopathic thrombocytopenic purpura to rule out leukemia? Arch Pediatr Adolesc Med 1998;152:345-7. Ovid Full Text Bibliographic Links [Context Link]
15. Gurwitch KD, Goldwire MA, Baker CJ. Intravenous immune globulin shortage: Experience at a large children's hospital. Pediatrics 1998;133:645-7. [Context Link]
16. Barnett V. Sample survey principles and methods. London, Edward Arnold, 1991. [Context Link]
Key Words: Thrombocytopenia; Purpura; Idiopathic thrombocytopenic purpura; Pediatric hematology/oncology; Questionnaire
Accession Number: 00043426-200001000-00011