The use of antibiotics in treating moderate paediatric community-acquired pneumonia (CAP). The antibiotics examined were intravenous (IV) ampicillin monotherapy, given at a dose of 100 mg/kg per day divided every 6 hours, and combination therapy with IV ampicillin followed by 5 mg/kg gentamicin given once daily.

Treatment.

The study sought to compare the cost-effectiveness of IV ampicillin monotherapy with IV ampicillin plus IV gentamicin in the treatment of paediatric CAP. The authors did not state the justification for their choice of the comparators. They also did not explicitly state a perspective for the study, but it can be inferred that it was conducted from a hospital perspective.

Cost-effectiveness analysis.

The study population comprised children aged between 2 months and 5 years, who were admitted to the paediatric ward and diagnosed with moderate CAP according to World Health Organization criteria. The diagnosis had to be confirmed by the attending physician and supported with a chest X-ray. Patients were excluded if they had any other medical or infectious problems, had received any medical intervention prior to admission, or were considered at risk of drug toxicity.

The setting was secondary care. The economic study was carried out in Malaysia.

The effectiveness and resource use data related to November 1999 to January 2000. The authors did not specify the source of the price data or the price year.

The effectiveness data were derived from a single study.

The cost data were collected prospectively from the same patient sample that was used in the effectiveness study.

The authors did not report any power calculations. Consecutive patients admitted to the paediatric ward were assessed for inclusion into the study. The authors did not state whether the initial study sample was appropriate for the study question. The total sample size was 40 patients, of which 20 were assigned to each treatment group. Out of 58 eligible patients, 40 were included in the study. The reasons for exclusions were not provided.

The study was a single-centred, randomised controlled trial. The authors did not describe the method of randomisation. The patients were followed up for the duration of their admission. No loss to follow-up was reported. The authors stated that the study was single-blind, but they did not state who was blinded to treatment allocation or how successful this was.

The analysis of effectiveness appears to have been conducted in an intention to treat basis. The primary health outcomes were:

improvement in signs and symptoms of pneumonia;

fever clearance time, defined as the time from the start of treatment until the auxiliary temperature fell below 37.5 degrees C and remained below that level for 48 hours;

the length of hospital stay;

the duration of IV antibiotic usage in the ward, defined as the time taken to switch from IV to oral antibiotic; and

antibiotic prescribed on discharge.

There was a significant difference between the groups in terms of the baseline sodium and potassium levels, (p<0.05). It was unclear whether there was any adjustment for confounding factors. The groups were shown to be comparable in terms of their age and weight on admission.

Fever clearance time was significantly shorter in patients receiving IV ampicillin monotherapy than in those receiving combination therapy, (p<0.05).

Patients receiving ampicillin monotherapy took a significantly shorter time to switch to oral antibiotic and had a significantly shorter time on IV antibiotic than those receiving combination therapy, (p<0.05).

The authors concluded that ampicillin alone gives better or comparable results in comparison with combination therapy.

No summary measure of health benefit was used in the economic analysis. The study was, in effect, a cost-consequences analysis.

The study included the direct hospital costs. These comprised drug acquisition and administration costs, laboratory costs, length of inpatient stay, and staff costs. The resource use quantities were not reported separately from the costs and this may limit the generalisability of the study results. The resource use data were based on the same patient sample as that used in the effectiveness study. The authors did not specify what unit cost or price data they used. Discounting was not relevant as the follow-up period was less than 1 year. The study reported the average cost

The indirect costs were not included in the analysis. This was appropriate for a study conducted from a hospital perspective.

Malaysian ringgit (MYR). The authors provide a conversion rate to US dollars ($), MYR 3.80 = $1.

The costs were analysed in a stochastic manner, but the authors did not specify the statistical test used. The authors compared the mean cost for each included cost item and the mean total cost between groups. No power calculations for the cost data were reported.

A sensitivity analysis was not carried out as the authors had access to individual patient level data and could thus use a statistical analysis.

See the 'Effectiveness Results' section.

The average total cost was MYR 143.55 (+/- 29.56) for patients receiving ampicillin monotherapy and MYR 222.86 (+/- 18.27) for patients receiving combination therapy. (The authors did not state whether the range provided was the standard deviation.)

The costs were calculated for the duration of admission. Discounting was not conducted, nor was it relevant.

The costs were significantly lower for patients receiving ampicillin monotherapy than for those receiving combination therapy, (p<0.05).

The authors did not report that any costs of adverse effects or knock-on costs were included.

The authors calculated average cost-effectiveness ratios for each primary health outcome measure. No summary measure of cost-effectiveness was provided.

The addition of gentamicin to ampicillin is not cost-effective. Ampicillin monotherapy is the most appropriate therapy for moderate paediatric community-acquired pneumonia (CAP).

The authors did not provide any justification for their choice of the comparators. You should decide whether these comparators are relevant in your own setting.

A prospective randomised controlled trial is an appropriate design for the study question. However, the authors did not specify who was blind to the treatment allocation. The groups were provided with other medication for pneumonia in addition to IV antibiotics, but the authors did not report whether there was any significant differences between the two groups in terms of these other medications. The authors did not state whether the study sample was representative of the study population. They mentioned that there were more males (55%) in the study sample than females. In addition, the majority of pneumonia cases were likely to have been caused by Gram-negative organisms, which may favour treatment with ampicillin monotherapy. There were some significant differences between the patient groups at baseline, and it was unclear whether these confounding factors were accounted for in the analysis. The authors did not specify which statistical tests were used for the effectiveness analysis. This makes it difficult to judge the credibility of the effectiveness results.

The authors did not derive a single measure of health benefits. The reader is referred to the comments in the 'Validity of estimate of measure of effectiveness' field (above).

The authors did not explicitly state a perspective for the cost analysis. The study included costs incurred during admission, and so excluded clinical events or adverse events post discharge. It is unclear whether this would have affected the authors' conclusions. The costs were not reported separately from the quantities. In addition, the source of the price data was not provided, nor was the price year. This reduces the generalisability of the study results. Although a statistical analysis was performed for differences in the mean costs, the type of statistical test used was not reported, which makes it difficult to judge the validity of the cost results.

The authors acknowledged that the short duration of the study meant that some potentially relevant outcome measures, such as antibiotic resistance, could not be measured. They also acknowledged that the treatment patterns in the study might not reflect current practice because culture and sensitivity tests were not conducted prior to antibiotic administration.

The authors did not make any recommendations for further research.

None stated.

This record was compiled by CRD commissioned reviewers according to a set of guidelines developed in collaboration with a group of leading health economists.