Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Heparin-Induced Thrombocytopenia-Associated Thrombosis in Pediatric Intensive Care Patients.

Schmugge M, Risch L, Huber A, Fischer JE.

Pediatrics 2002; 109:e10. [abstract]

 

Reviewed by Kshitij P. Mistry, MD, Children's Hospital, Boston

Review posted April 2, 2003

I. What is being studied?:

The study objective:

To determine the incidence of HIT-associated thrombosis in heparin-exposed children.

The study design:

Reported as a retrospective cohort study design (additionally, case-control element was utilized to compare antibody levels; see below)

The patients investigated:

All patients admitted between August 1996 and January 1999 to a 19-bed multidisciplinary pediatric intensive care unit serving as the tertiary referral center for eastern Switzerland. Patients (n = 612) who received heparin for > five days were eligible.

II. Are the results of the study valid?

Primary questions:

1.Were there clearly identified comparison groups that were similar with respect to important determinants of outcome, other than the one of interest?

The authors suggest that the study design is a retrospective cohort study in which an investigator identifies exposed and non-exposed groups of patients and subsequently ascertains the occurrence of the outcome in both patient populations.

Upon review of the methodology, it appears that this research study may more closely resemble a case-control design. In this method, the investigator identifies patients in whom the outcome has already occurred (in this case, thrombosis) and then assesses retrospectively the frequency of exposure (clinical and radiological confirmation of heparin-induced thrombocytopenia). No controls (patients without the outcome of interest, i.e. thrombosis) were chosen to compare frequency of HIT "exposure."

The authors selected two controls for each case (matched for age, diagnosis, intervention, duration of heparin administration and sample storage time) to determine the appropriate ELISA HPF4 antibody level "cutoff," suggesting laboratory confirmation of HIT.

2. Were the exposures and outcomes measured in the same way in the groups being compared?

In case-control studies, ascertainment of the exposure is a key issue.

To be included in this study, patients must have been exposed to ≥ five days of heparin therapy.

The diagnosis of heparin-induced thrombocytopenia was based on clinical HIT criteria (thrombocytopenia [platelet count below 150 X 10⁹/L] or decrease in the platelet count of 50% occurring at least five days after heparin exposure) as well as serologic confirmation (ELISA HPF4 antibody levels). Since the latter evaluation is primarily utilized in adult patients, no threshold for pediatric patients exists. As a result, the authors created an arbitrary level based on comparison with control subjects.

Additionally, 38 of the 57 patients with thrombosis (cases) were not included because no plasma samples were available to perform the ELISA HIT confirmatory test.

The authors identified patients with thrombosis as those with clinical symptoms as well as radiological confirmation (Doppler-ultrasonography, angiography or computed tomography scan). No definition of "clinical symptoms" was provided.

Again, no specific control group was identified for comparison. In particular, we are not told the incidence of thrombosis in those not exposed to heparin.

3. Was follow-up sufficiently long and complete?

Yes. Thrombosis was diagnosed after a median of ten days (range of 5-45 days) following initiation of heparin administration. Serologic studies suggest that heparin-induced antibodies usually develop five to eight days following heparin exposure and immune-mediated HIT occurs four to ten days after commencement of heparin therapy (1).

Secondary questions:

4. Is the temporal relationship correct?

Yes. Thrombosis was diagnosed after a median of ten days (range of 5-45 days) following initiation of heparin administration. This suggests exposure to heparin (and subsequent development of HIT) preceded thrombosis formation.

5. Is there a dose-response gradient?

Uncertain. Patients received unfractionated porcine heparin intravenously whenever central venous access was obtained, arterial lines were placed or for prophylaxis and treatment of thrombosis. The authors do not provide dosing or length of treatment for these regimens; however, six of the 14 patients diagnosed with HIT had been exposed to heparin on previous hospitalizations and 11 of the 14 had undergone cardiac surgery (with intraoperative exposure to heparin). Additionally, we do not know the previous exposure or length of heparin therapy of patients without HIT-associated thrombosis.

III. What are the results?

1. How strong is the association between exposure and outcome?

Thrombosis	Yes	No
HIT	14	B
No HIT	24	D
Totals	38*	555	612

* 19 patients with thrombosis were not included secondary to unavailability of plasma for ELISA evaluation of the HPF4 antibody level.

In this study, the number of patients exposed to heparin therapy with and without HIT in whom thrombosis did not develop (cells B and D) is not provided and therefore, the appropriate measure of association (the odds ratio [OR]) cannot be calculated.

We can, however, calculate the incidence of thrombosis in the cohort aspect of this study; 57 patients of 612, or 9%, of those exposed to heparin developed thrombosis. Of those with plasma to sample, HIT was diagnosed in 37% (14 of 38) of patients with thrombosis exposed to heparin. Without a comparison group of those not exposed to heparin, we do not know whether heparin represents a risk factor for thrombosis. To be rigorous, from this data we cannot truly ascertain whether exposure to heparin (as arbitrarily defined here) is a risk factor for HIT. We can only conclude that in patients with thrombosis exposed to heparin, there appears to be a high rate of associated HIT.

2. How precise is the estimate of the risk?

The precision of a measure of association is evaluated by the confidence interval around the estimate. In this study, the precision cannot be assessed because the odds ratio cannot be calculated with the data provided.

IV. What are the implications for my practice?

1. Are the results applicable to my practice?

The patients included in this study represent "the entire spectrum of pediatric critical illness and included patients who had undergone cardiac and other major surgery" (2). Although the specific diagnoses are not provided, this general description suggests a comparable patient population to other tertiary care, multidisciplinary pediatric intensive care units. However, many PICUs do not care for cardiac ICU patients, which comprised 11 of the 14 children found to have HIT-associated thrombosis.

2. What is the magnitude of the risk?

The authors suggest the observed incidence of HIT-associated thrombosis was 2.3%, with a 95% confidence interval of (1.3%, 3.9%). Additionally, they conclude that the proportion of thrombotic events attributed to heparin-induced thrombocytopenia in patients receiving at least five days of heparin therapy is 37%, with a 95% confidence interval of (22%, 54%).

As mentioned previously, it is difficult to interpret the latter assumption since individuals with HIT and absence of thrombosis were not included (table 1). 19 children with thrombosis were not included in the study secondary to unavailability of plasma samples. Therefore, the proportion of thrombotic events attributed to heparin-induced thrombocytopenia may be as high as 57.9% (33/57 patients with thrombosis).

Furthermore, the 2.3% incidence value is based on 14 of 38 patients with thrombosis having HIT; as mentioned previously, 19 children with thrombosis were not included in the study secondary to unavailability of plasma samples. Therefore, the incidence may be as high as 5.4% (33/612 total patients).

To determine the number of people that must be exposed to cause an adverse outcome, we would need the absolute risk of thrombosis development associated with HIT and absence of HIT.

3. Should I attempt to stop the exposure?

The authors report a significant increase in the magnitude of the risk of developing NSeveral aspects must be considered when evaluating a potential harmful exposure: the strength of inference, magnitude of risk, adverse consequences of reducing the exposure and available alternatives to therapy (3). This study suggests that in the pediatric patient population HIT-associated thrombosis occurs more often in extrapulmonary sites, in children less than one year of age and equally within the arterial and venous circulation. The magnitude of this risk, however, remains unclear and a prospective study is needed to accurately determine the incidence.

Several alternatives to unfractionated heparin exist, including utilization of low molecular weight heparin and warfarin therapy (3,4,5). Limiting heparin to less than five days may not be clinically practical; therefore, vigilance of clinical symptoms and laboratory evidence of heparin-induced thrombocytopenia may be the best regimen.

References

1. Coutre S. Heparin-Induced Thrombocytopenia. UpToDate®. www.uptodate.com. 2002.
2. Schmugge M, Risch L, Huber A, Fischer JE. Heparin-Induced Thrombocytopenia-Associated Thrombosis in Pediatric Intensive Care Patients. Pediatrics 2002; 109:e10. [abstract]
3. Levine M, Walter S, Lee H, Haines T, Holbrook A, Moyer V. Users' guides to the medical literature. IV. How to use an article about harm. Evidence-Based Medicine Working Group. JAMA 1994; 271(20):1615-9. [full-text]
4. Valentine KA, Hull RD. Clinical Use of Heparin and Low Molecular Weight Heparin. UpToDate®. www.uptodate.com. 2002.
5. Mannucci PM. Drug Therapy: Hemostatic Drugs. NEJM 1998; 339(4):245-253. [link]

Comments

Back to the EB Journal Club Index