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

Inhaled nitric oxide in full term and nearly full term infants with hypoxic respiratory failure.

The Neonatal Inhaled Nitric Oxide Group

N Engl J Med 1997;336:597-604. [abstact]

Reviewed by P. Chhangani, MD and J. Hupert, MD, University of Illinois at Chicago

Review posted June 19, 1999

---

I. What is being studied?:

The study objective:

To determine whether inhaled nitric oxide would reduce mortality or need for ECMO in term or near-term infants with hypoxic respiratory failure unresponsive to conventional aggressive therapy.

The study design:

Multicenter, prospective, randomized, double-blind, placebo-controlled clinical trial

The patients included:

• Infants older than or equal to 34 weeks gestational age
• Age equal to or less than 14 days
• Oxygenation index of > 25 on two measurements at least 15 minutes apart Presence of indwelling catheter
• Undergo Echocardiography prior to randomization

The patients excluded:

• Infants older than 14 days
• Presence of congenital heart disease or congenital diaphragmatic hernia
• Decision not to provide full treatment

The interventions compared:

The infants were randomized to receive either 100% O2 (control) or 20 ppm of NO after the second qualifying blood gas. After 30 mins of initiation of study gas (either NO or O2), paO2 was measured, responses were:

• complete: if paO2 increased by 20 mmHg
• partial: if paO2 increased by 10-20 mmHg
• none: if paO2 increased by < 10mmHg

If the response was complete then study gas continued. In cases of partial or no responses then the study gas was stopped for 15 mins and restarted at 80 ppm. Responses measured again after 30 mins and treatment continued at that concentration if response was complete, continued at the lowest concentration that had produced a partial response in cases of partial responses and discontinued in cases of no response.

The outcomes evaluated:

Primary Outcome: Death or initiation of ECMO by 120 days.

Secondary Outcome: Whether NO would increase paO2 and decrease OI and AaDO2 thirty minutes after the start of treatment. The authors sought to determine if this response would reduce hospitalization without an increase in duration of assisted ventilation or incidence of air leak or bronchopulmonary dysplasia (BPD).

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

Yes, the patients were stratified according to center and then randomly assigned by telephone to receive either 100% O2 or NO by permuted block design in the center.

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Was followup complete?

Two hundred and thirty five patients were enrolled and all follow up was complete.

Were patients analyzed in the groups to which they were randomized?

Analysis of primary outcome (death or initiation of ECMO) was by "intention to treat" and there were no crossovers. In the physiological responses to gas administration, six infants were excluded (4 controls and 1 NO patient did not receive any gas, and one NO patient's data was unavailable for technical reasons). Seven patients who received the wrong study gas (6 controls and 1 NO) were included in the analysis based on assigned treatment.

Secondary questions:

3. Were patients, health workers, and study personnel "blind" to treatment?

Every effort was made to keep the clincal teams "blind," initially by mock adjustments in control gas. However after the initial dosing algorithm the gas was monitored by study individuals (not involved in clinical care) in an unmasked fashion. The authors then state that decisions about ECMO were made by a blinded clinical team on the basis of center specific criteria. It is unclear from the paper how further monitoring could be unmasked yet the clincal team remained blinded.

4. Were the groups similar at the start of the trial?

Yes, the groups had similar baseline characteristics, e.g., birth weight, gestational age, M:F ratio and the clincal status at the time of randomization. This status includes the type of conservative treatment, air leaks, pulmonary hemorrhage, parameters on qualifying blood gas and age at randomiaztion.

5. Aside from the experimental intervention, were the groups treated equally?

Each participating center developed general management guidelines to be used before and throughout the treatment; these included maintainance of mean arterial blood pressure > 45 mmHg, induction of alkalosis (pH 7.45 to 7.6) and the use of surfactant.

III. What were the results?

1. How large was the treatment effect?

a) Primary Outcome (decrease in mortality rate or rate of initiation of ECMO):

Outcomes	Control (n=121) X	NO (n=114) Y	ARR (X-Y)	RRR (X-Y/X)	NNT (1/X-Y)
Death/ECMO	77(64%)	52(46%)	18%	28%	6(3,20)
ECMO	66(55%)	44(39%)	16%	29%	6(3,33)

Thus to prevent death/ECMO in 1 patient we would have to treat 6 patients. The incidence of death in both groups was not statistically significant as evidenced by the p value of > 0.6 which means that the difference could occur by chance 60% of the time.

b) Secondary Outcome (increase in paO2 30 mins after initiation of study gases):

Outcomes	Control(n=117)	NO(n=112)	ARR	NNT(95% C.I)
None	87(75%)	38(34%)	41%	2(2,3)
Combined(Partial and Complete)	30(26%)	74(66%)	40%	3(2,4)

Thus it is seen that we would need to treat 3 babies to achieve an increase in paO2. There were no significant differences amongst survivors between the groups in the rates of airleak syndrome or BPD, and there was no difference in the duration of assisted ventilation or hospitalization.

2. How precise was the estimate of the treatment effect?

a) Primary Outcome: As shown in the table above, the 95% confidence intervals of 3 and 20 around the NNT in the category of initiation of ECMO, we would have to treat as few as 3 or as many as 20 babies to prevent one case of initiation of ECMO.

b) Secondary Outcome: It is apparent from the confidence intervals of the NNT of 3 (2,4) that the effect would be seen when as few as 2 or as many as 4 babies would be treated.

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

1. Can the results be applied to my patient care?

Yes, the spectrum of patients (especially as this was a multicenter study), applicability of inclusion criteria, availability and applicability of method of treatment all help in deciding to use NO in infants with hypoxic respiratory failure. The significant difference in the rate of initiation of ECMO in the NO group would also help in the decision. In particular, it appears that babies with a diagnosis of pulmonary hypertension have a significantly reduced risk of primary outcome (death or ECMO) with NO compared to controls; RR= 0.43 (95% CI 0.23,0.81). Infants with a diagnosis of pneumonia or sepsis also appear to benefit; RR = 0.58 (95% CI 0.33,1.00)

2. Were all clinically important outcomes considered?

Yes, reduction in rates of initiation of ECMO or death seem to be appropriately important outcomes. Complications in survivors, such as rates of air leak syndrome and BPD, as well as duration of assisted ventilation and hospital stay, were also evaluated.

3. Are the likely treatment benefits worth the potential harms and costs?

Probably. Since initiation of ECMO is an invasive and potentially hazardous intervention and it was also seen in this study that NO did not have any detectable clinical side effects; the study drug was not discontinued due to toxic effects as seen by methemoglobin and NO2 levels. However the number of patients was not large enough to demonstrate even a 10% increase in the occurrence of IVH.

Comments

Back to the EB Journal Club Index