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

Outcome of children who require mechanical ventilatory support after bone marrow transplantation.

Keenan HT, Bratton SL, Martin LD, Crawford SW, Weiss NS.

Crit Care Med. 2000;28(3):830-5. [abstract]

Reviewed by Xiomara Garcia, MD, M. Hossein Tcharmtchi, MD. Section of Critical Care Medicine, Department of Pediatrics, Baylor College of Medicine, Houston, Texas.

Review posted November 12, 2000

I. What is being studied?

Study objective:

To identify clinically measurable factors that could predict outcome for pediatric patients undergoing mechanical ventilatory support after bone marrow transplantation.

Study design

A retrospective cohort study. Patients included: all children < 17 yrs old who received a bone marrow transplant and later required mechanical ventilatory support for > 24 hours at the Fred Hutchinson Cancer Research Center (FHCRC) between 1983 and 1996. The FHCRC is a referral center in Seattle, Washington for bone marrow transplant patients. Patients were identified by international classification of diseases ICD-9 codes for mechanical ventilatory support.

Patients were excluded if they received mechanical ventilatory support for less than 24 hours, received only continuous positive airway pressure, or if they required transfer for high frequency oscillatory ventilation or subspecialty care at another hospital.

II. Are the results of the study valid?

Note: These questions follow from Randolph AG et al. Understanding articles describing clinical prediction tools. Crit Care Med 1998;26:1603-1612. [abstract]

1. Was a representative group of patients completely followed up?

Yes. The medical records of 1080 pediatric patients with bone marrow transplantation were reviewed at the FHCRC, between 1984 to 1996. There were 152 patients who later received mechanical ventilatory support. Twenty six were excluded for being ventilated less than 24 hours, and 5 were excluded for being transferred to Children's Hospital and Medical Center for high frequency oscillatory ventilation. All five of these patients died and perhaps should ideally be included in the study cohort. The remaining patients were followed for the duration of the study.

Was a follow up sufficiently long and complete?

Yes, apparently. Three patients transplanted before 1984 who received mechanical ventilation in 1984 were included in the cohort. The maximum follow up period was through July of 1997; therefore there is no information about delayed need for ventilation in patients who were transplanted during the later years of the period studied.

Only 19 patients survived for more than 30 days after extubation. Twenty eight days is accepted as a primary outcome of mortality studies in the ICU.

2. Were all potential predictors included?

Most important predictors were included.

Data abstracted included demographic information, indication for transplant, and disease status. Characteristics of the patient's transplant, including the conditioning protocol (chemotherapeutic agents, dose of irradiation), type of transplant (allogeneic, matched, mismatched, or autologous), and success of engraftment (grafted versus nonengrafted) were reported. Patients were identified by international classification of diseases ICD-9 for mechanical ventilatory support. Indication for endotracheal intubation were: procedure/surgery, airway protection secondary to mucositis or decreased mental status, respiratory failure and cardiopulmonary arrest. Ventilatory support settings and radiology reports were recorded for each 24 hour period. Laboratory data were collected for every 24 hour period of ventilation. Worst values for liver enzymes, creatinine and blood urea nitrogen were recorded. Positive cultures from any site were noted. Additional data collected included presence or absence of GVHD (Seattle grading system), hepatic veno-occlusive disease (VOD) and need for hemodialysis. Serum chemistry values indicative of organ dysfunction were defined on the basis of Nichols et al.(1).

3. Did the investigators test the independent contribution of each predictor variable?

Yes, the authors tested the independent contribution of each risk factor (ARDS, renal insufficiency, hepatic insufficiency, use of vasopressors) for survival. The combination of risk factors for survival at each time period (day of intubation, day 2 and day 7 postintubation) was also analyzed. The authors did not enter the predictor variables into a logistic regression model.

Fortunately determining whether a tool is clinically useful has little to do with the mathematical method used to build the tool. It is more important to confirm that variables were not only considered individually, but together (2).

4. Were outcome variables clearly and objectively defined?

Yes, all variables were clearly and objectively identified, except for the use of vasopressors. The parameters used to initiate or discontinue these drugs were not clearly defined.

Ironically, one outcome measure was not precisely defined - death. There is no mention of limitation of care or withdrawal of support. It would be very important to know which patients, if any, died from removal of ventilatory support and which died despite continued escalation of care.

III. What are the results?

1. What is(are) the prediction tool(s)?

The authors identified variables associated with mortality and mechanical ventilatory support after bone marrow transplantation, but these variables were not arranged into a model that allows the prediction of risk of mortality.

The study was successful in defining factors that decreased risk of survival, but could not sufficiently predict survivors from non-survivors to provide clear clinical guidance.

There were no appreciable differences between survivors and non-survivors by variables such as age, gender, race, characteristics of the transplant (initial diagnosis, type of transplant), and GVHD. Patients with VOD and those who underwent dialysis had a moderately reduced chance of survival. The one predictor strongly associated with risk of death was indication for intubation secondary to respiratory failure (RR: 0.119; 95% CI 0.028, 0.396) and pulmonary infection (RR survival: 0.2; 95% CI: 0.07-0.7). All the relative risks are, paradoxically, the "risk" of survival.

The only factor that was strongly associated with reduced survival on the day of intubation was the presence of ARDS (RR: 0.3; 95% CI 0.1, 0.8). Hepatic insufficiency, renal insufficiency, and vasopressor use were not by themselves associated with survival status. On both days 2 and 7 postintubation, the presence of either hepatic insufficiency or ARDS was associated with death. On postintubation day 7, patients who required vasopressors (RR survival: 0; 95% CI : 0.0, 6.8) or had renal insufficiency (RR survival: 0; 95% CI: 0.0, 4.1) did not survive.

Combinations of risk factors for survival at each time interval were examined including ARDS combined with renal and/or hepatic dysfunction or with the use of vasopressors. No patient who had the combination of ARDS and renal insufficiency at the time of intubation survived. On day 2 the combination of ARDS and renal insufficiency remained strongly associated with death as was the combination of ARDS and hepatic insufficiency. On day 7 postintubation, any combination of organ insufficiencies was associated with a high chance of death (there was one survivor with ARDS + hepatic insufficiency); RR :0.1; 95% CI: 0.02, 1.0)

2. How well does the model categorize patients into different levels of risk?

The authors did not develop a formal model to categorize patients into different subgroups of risk, i,e,, raw mortality rates are not specified.

This study does provide a set of prognostic factors that clinicians can use when considering whether or not to continue invasive medical therapies for high risk patients.

In summary, high risk patients would include those with: (relative risk is for death, compared to individuals in the cohort without this risk factor)

• Intubation secondary to respiratory failure (RR 0.12, 95% CI 0.03, 0.40)
• Pulmonary infection (RR 0.2, 95% CI 0.07, 0.7)
• ARDS on day of intubation or days 2 or 7 (RR 0.1 to 0.3, upper 95% CI always < 0.7)
• Hepatic insufficiency on day 2 and 7 ( RR: 0.2 - 0.3, upper 95% CI < 0.8)
• Multiple organ failure (only ARDS + hepatic insufficiency on day 2 was significant; RR 0.3 95% CI 0.1, 0.8)

3. How confident are you in the estimates of risk?

Although the size of the study group was more than twice the size of any previous pediatric group, the number of survivors was low and confidence intervals were wide. Therefore the predictive value of the risk factors for children are not certain.

IV. Will the results help me in caring for my patients?

1. Does the tool maintain its prediction power in a new sample of patients?

No. The authors did not develop a prediction tool. In order to validate the predictions offered, a separate sample of patients would be required.

2. Are your patients similar to those patients used in deriving and validating the tool(s)?

Yes, the study patients were representative of typical bone marrow transplant patients who require PICU care. The Fred Hutchinson Cancer Research Center (FHCRC) is a referral center for bone marrow transplant patients. They may accept higher risk patients than other transplant programs.. At Texas Children's Hospital our bone marrow transplant patients are comparable with FHCRC. The mortality rate in this group of patients is high. In the past year, in our experience patients that were intubated secondary to mental status changes, pulmonary hemorrhage or allergic reaction (e.g., to amphotericin B lipid complex) with airway obstruction and pulmonary edema survived even after 7 days of mechanical ventilatory support (n= 4).

The trend toward increasing mortality associated with increasing multiple organ system failure was demonstrated by Wilkinson et al. in 1986(3). Rossi confirmed the same results in 1999, but their study also showed an increase in survival rates of children requiring mechanical ventilatory support after bone marrow transplantation, although the reason was unclear(4). Though Denardo argued that mechanical ventilation for > 4 days in adults is futile (5), there are not well defined studies that specify a time limit of mechanical ventilation in pediatric population.

3. Will the results lead directly to selecting or avoiding therapy?

No. The present study was successful in identifying the risk factors for decreased survival. The study could not sufficiently distinguish survivors from non-survivors with enough certainty to provide clear clinical guidance.

Multiple organ failure syndrome and respiratory failure in children with bone marrow transplantation is not sufficiently discriminative to determine continuation or withdrawal of ICU support.

4. Are the results useful for reassuring or counseling patients?

Once again, even though the study could not predict survivors from non-survivors, this study did provide a set of prognostic factors than can be discussed with patients. Families and physicians can then decide whether or not it is reasonable to continue invasive medical therapies, while recognizing that some possibility of survival may be very small.

References

1. Nichols DG, Walker LK, Wingard JR, et al. Predictors of acute respiratory failure after bone marrow transplantation in children. Crit Care Med. 1994;22(9):1485-91. [abstract]
2. Randolph AG, Guyatt GH, Calvin JE, et al. Understanding articles describing clinical prediction tools. Crit Care Med 1998;26:1603-1612. [abstract]
3. Wilkinson JD, Pollack MM, Ruttimann UE, Glass NL, Yeh TS. Outcome of pediatric patients with multiple organ system failure. Crit Care Med. 1986;14(4):271-4. [abstract]
4. Rossi R, Shemie SD, Calderwood S. Prognosis of pediatric bone marrow transplant recipients requiring mechanical ventilation. Crit Care Med. 1999;27(6):1181-6. [abstract]
5. Denardo SJ, Oye RK, Bellamy PE. Efficacy of intensive care for bone marrow transplant patients with respiratory failure. Crit Care Med. 1989;17(1):4-6. [abstract]

 

Comments

Back to the EB Journal Club Index