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

Chest physiotherapy for the prevention of ventilator-associated pneumonia

Ntoumenopoulos G, Presneill JJ, McElholum M, Cade JF.

Intensive Care Med 2002;28: 850-856. [abstract; full-text for subscribers]

Reviewed by Nancy Hoover MD, and Jana Stockwell MD Pediatric Critical Care Children's Healthcare of Atlanta

Review posted January 17, 2003

---

I. What is being studied?:

The study objective:

To determine whether chest physiotherapy (CPT) decreases the occurrence of ventilator-associated pneumonia (VAP).

The study design:

Prospective controlled systematic allocation trial.

The patients included:

Consecutive general medical, surgical and/or trauma adult patients admitted to a University teaching hospital ICU during a 6 month period (September 1997-February 1998). Patients were eligible for study enrollment if they required oral intubation and mechanical ventilation for at least 48 hours.

Over the 6-month trial period, 60 out of 253 intubated and mechanically ventilated patients admitted to the ICU were enrolled.

The patients excluded:

177 patients were excluded from the study secondary to predetermined exclusion criteria such as open heart surgery within 24 hours prior to admission, aspiration or community-acquired pneumonia, mechanical ventilation <48 hours, leukopenia (WBC count < 4,000), or if the patient had sustained any injury or developed any complication that did not allow chest physiotherapy to be provided. A further 16 patients were removed subsequently due to death less than 48 hours after admission to the ICU, 6 were from the intervention group and 10 from the control group, so 60 (60/253; 24%) patients fulfilled all the entry criteria.

The interventions compared:

Patients allocated to the intervention group received CPT twice-daily beginning on the day of admission to the ICU. CPT comprised gravity-assisted drainage or positioning on side for 20 minutes with affected lung up, four sets of 6 cycles of expiratory chest wall vibrations and airway suctioning (at least 3 times) via ETT or tracheostomy tube interspersed through the treatment.

Once the patient was extubated or removed from the ventilator, CPT comprised of the same routine of patient positioning and expiratory chest wall vibrations, combined with coughing or tracheostomy suctioning.

The control group did not receive CPT while in the ICU, however "sham" treatment was provided by the therapists.

ICU nursing staff, without limit as required by the patient's condition, could provide patient re-positioning side-to-side and airway suctioning.

The outcomes evaluated:

Primary outcomes were the development of VAP. Secondary outcomes of interest included: duration of mechanical ventilation, median length of stay in the ICU and the mortality at 28 days.

The data to determine the clinical pulmonary infection score (CPIS, see Table 1 in the paper) were prospectively obtained daily by the chief investigator. As part of daily routine ICU care, all patients were clinically evaluated for signs of pneumonia, including CXR and routine clinical parameters. Each patient was then rated by the attending ICU physician as either having or not having a clinically suspected pneumonia.

VAP was defined as being present when, on the same calendar day, both the clinical diagnosis and total CPIS score indicated the presence of pneumonia. VAP was defined as early onset when it occurred < 5 days and late onset (> 5 days) after ICU admission.

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

No. The patients were systematically assigned based on day of admission to the ICU rather than in a blind, randomized allocation. There were 24 patients allocated to the intervention group and 36 allocated to the control group. Great bias could be introduced by this lack of true randomization. All the patients began receiving CPT or not on day one of admission according to the protocol. Those that stayed longer than 48 hr were then followed and continued on the same treatment regimen for their entire ICU stay.

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

Was followup complete?

Yes. The 16 patients who died within 48 hours after admission to the ICU were noted above - it appears they were entered into the study and, by the intention-to-treat principle, perhaps should have been included in the analysis.

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

The patients were analyzed in the groups they were assigned to, but again, the patients were not truly, randomized to the treatment group or the control group.

Secondary questions:

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

Blinding of the ICU medical staff and nursing staff was done with "sham" physiotherapy in the control group. The authors state that the clinical staff was unaware of the prospectively defined outcome of VAP. Clearly the physicians knew that a patient developed VAP, but it seems that they may not have known it was being analyzed and compared with the CPIS score being done independently.

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

Yes. There was no significant difference between the 2 groups in demographic The two trial groups were well matched for age, sex, admission APACHE II and GCS scores and P/F ratio. During the ICU stay, there were no clinically significant differences in the frequency of cardiac arrest, use or IV narcotics, sedatives, paralytics or ICP monitors/drains. In the surgical subgroup, those in the control group had a significantly higher ASA score.

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

Yes. It is assumed that they all received the same standard of care according to their ICU guideline. No mention of ranitidine or sulcralfate treatment was provided. The amount and degree of suctioning by the nurse after CPT or sham CPT may have been different, but it is not reported. Standard medical and nursing care was provided for both groups according to their ICU guidelines. Early enteral nutrition was encouraged in both groups and no patient received selective decontamination of the digestive tract. (The enteral route, gastric or small bowel, was not specified.) Other medical care provided as clinically indicated included antibiotic therapy and bronchoscopy. Tracheostomy was performed as per ICU policy.

III. What were the results?

1. How large was the treatment effect?

They demonstrated a significant reduction in VAP from the control group 14 (39%) compared to the intervention group 2 (8%) with an OR of 0.14, 95% confidence interval of 0.03-0.56. It is not clear why, in this prospective trial, they are using the odds ratio; a more appropriate metric is the relative risk, which in this case is 0.2, with 95% CI's of 0.05 and 0.86.

Those patients with higher APACHE II scores (17 in the control group and 13 in the intervention group) had a lower frequency of VAP in the intervention group (n = 0) compared to the control group (n = 5) with an absolute risk reduction of 0.29 and 95% CI's of 0.04 and 0.54.

Patients ventilated for longer than the overall median of 4.65 days, had a statistically significant reduction in the frequency of VAP in the intervention group (n = 1) compared with the control group (n = 11) with on odds ratio of 0.09, 95% CI 0.11, 0.79. This is curious; the RR is 0.18 with 95% CI's that cross one; 0.03, 1.22. The numbers are quite small and they used the Fisher Exact test for their p value - perhaps this explains the discrepancy.

Multivariate analysis demonstrated that CPT remained independently associated with a reduced frequency of VAP following adjustment for several factors including the presence of tracheostomy, the severity of illness and the duration of mechanical ventilation with odds ratio of 0.16, 95% CI 0.03, 0.94.

The significant increase VAP diagnoses in the control group was not coupled with longer time on ventilator or ICU stay and that group actually had improved 28 day mortality. Perhaps this is due in part to the small numbers of patients or maybe just increased atelectasis in the control group.

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

Precision of the treatment effect is demonstrated by the confidence intervals included.

The investigators assumed a VAP proportion of 40% in control patients and in order to achieve 80% power and 20% absolute reduction, they determined the study needed 80 patients in each arm of the study. They had only 45% of that in the control group (36/80) and 30% in the treatment group (24/80), but were still able to show a significant difference between the groups.

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

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

Maybe, but with some caution since my patient population is pediatric and there is a large difference in ET tube size that may lead to reduced secretion clearance despite aggressive CPT. In our Pediatric ICU, CPT is usually only ordered for patients with radiographic evidence of atelectasis. Patients do not routinely receive CPT secondary to the need for increased sedation. We tend to use ventilator strategies that promote the reversal of atelectasis such as longer inspiratory times and cycled prone positioning Based on this paper, I would consider a change in my practice to much more aggressive CPT in order to prevent VAP.

2. Were all clinically important outcomes considered?

Yes. The lack of difference in duration of ventilation or length of stay is not consistent with the body of literature that suggests VAP increases these outcomes or shows no benefit of CPT to reduce VAP (1,2). This makes the applicability of this study all the more questionable. The variable missing is the link with the nursing staff. The timing and aggressiveness of nursing suctioning may be a significant contributor to VAP.

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

The study did not discuss the increased cost of the CPT, which might be significant since it was not offset by a significant reduction in ICU stay, or mortality. The potential harms not discussed are the numbers of accidental extubations and the need for additional sedation. This is a problem, particularly in caring for children who are often unable to cooperate with their therapy.

References

1. Kollef MH. The prevention of ventilator-associated pneumonia. N Engl K Med 340;627-634, 1999. [abstract]
2. Hall JC, Tarala RA, Tapper J, Hall JL. Prevention of respiratory complications after abdominal surgery: a randomized clinical trial. Br Med J 312:148-152, 1996. [abstract]

Comments

Back to the EB Journal Club Index