Negative pressure wound therapy for wounds: as yet no convincing evidence of an additional benefit

There is currently no convincing evidence that wounds treated with negative pressure wound therapy (NPWT) heal better than those treated with conventional methods. Additional clinical trials should therefore be conducted and NPWT should not be widely employed for the time being. This is the main finding of a final report published by the Institute for Quality and Efficiency in Health Care (IQWiG) on 16 May 2006. However, IQWiG expects further studies to be completed in the near future and believes that there may then be stronger evidence available to support an additional benefit of NPWT. It therefore recommends that the method be re-evaluated in 2 to 3 years.

The German Federal Joint Committee ( Gemeinsamer Bundesausschuss, G-BA) commissioned IQWiG to evaluate the benefits and harms of NPWT compared with conventional therapy in patients with acute or chronic wounds, on the basis of the available literature. In Germany, NPWT has already been used in hospitals for several years in the hope that it can accelerate the healing of large wounds. In the inpatient sector, the statutory health insurance funds reimburse the costs of this treatment, but not the costs for its use in outpatient settings. An application has been submitted to the Federal Joint Committee for recognition of NPWT as a standard therapy in outpatient treatment and its reimbursement by the statutory health insurance funds.

Poor quality of the studies overall

In a comprehensive literature search, the scientists in Cologne identified 17 studies with a total of 602 participants in which NPWT was compared with conventional wound treatment and which provided at least relatively reliable results. However, the methodological quality of these studies was regarded as poor. The most frequent deficits included contradictory results even within a single study. In some studies, it was also unclear whether the data analysis had been conducted in an appropriate manner. For example, it is important that all patients who take part in a study are also included in the evaluation. Often it is precisely those patients who drop out of a study before the end (for example, due to adverse effects or complications) who are of particular importance for the evaluation. In many of the studies analysed by IQWiG it remained unclear how such drop-outs had been handled. Finally, even in cases where it would have been possible, the endpoints were often not investigated in a blinded design, which increases the risk of bias.

Seven of the 17 studies were RCTs (randomised controlled trials) in which the patients had been randomly assigned to 1 of the 2 groups. Ten non-randomised intervention studies, in which the patient groups at least received therapy at the same time, were also included in the evaluation.

No long-term results available

Although the manufacturers often claim differently, the studies investigated did not produce any convincing evidence that wounds heal faster with NPWT than with conventional therapy. While the results of 2 of the 5 randomised trials that investigated time to healing as an endpoint were favourable for NPWT, the 3 others failed to demonstrate significant differences. Also, only 2 out of 4 of the non-randomised trials showed that wounds healed faster with NPWT. Although surrogate parameters (the reduction of wound area and volume) provided indications for faster healing under NPWT, in the opinion of the Cologne scientists the studies were impaired by various methodological deficits and thus their conclusions are not really reliable.

The results on late complications, such as amputations and recurrent wounds, are also questionable. These results have little evidential value, partly because none of the patients in any of the studies were followed up for longer than a year after conclusion of the treatment. In most cases, the follow-up period was way below the minimum follow-up period of 3 months after wound closure stipulated by the US Food and Drug Administration. With regard to the avoidance of first-time or additional amputations, the results for NPWT were somewhat better in one of the larger studies. Since it is unclear whether the drop-outs were counted as a "therapy success", this result - which was in any case not statistically significant - is difficult to evaluate.

Quality of life and pain not investigated in the studies

It also still remains an open question whether patients whose wounds were treated with NPWT can be discharged faster from hospital or whether they are hospitalised less often. As far as avoiding hospitalisation is concerned, none of the studies included found a statistically significant difference between NPWT and conventional therapy. No conclusions can be drawn with regard to the length of hospital stay, due to methodological deficits: this endpoint was only investigated in non-randomised trials in which no blinding was performed.

According to the IQWiG scientists, the impact of NPWT on quality of life, pain during treatment, and scar formation is also important. Unfortunately these questions were either not investigated systematically in the studies evaluated or the relevant data were not included in the publications.

Conclusions on outpatient use of NPWT hardly possible

In summary, the IQWiG scientists describe the evidence base for the use of NPWT as meagre, at best. This applies both to the inpatient and in particular to the outpatient sector. Since the study data were collected mainly in inpatients, and outpatient NPWT was only partly used in 3 of the 17 studies included in the IQWiG review, conclusions as to a relevant additional benefit of NPWT in the outpatient sector are associated with even greater uncertainty and are thus not currently possible. However, the IQWiG scientists hope that this will change, as the results of several ongoing randomised trials will be published in the next few years. They therefore recommend that the evidence be re-evaluated in about 2 to 3 years.

Procedure of report production

On 21 December 2004, the Federal Joint Committee commissioned IQWiG to produce a report. This commission was specified 3 months later on 21 March 2005 in agreement with IQWiG. On 13 May 2005, IQWiG published a report plan and the preliminary report followed on 1 December 2005. A large number of statements were submitted on the preliminary report. These were reviewed and the most important aspects were discussed at length with those who submitted statements in a scientific hearing at IQWiG on 10 January 2006. The results of the hearing were incorporated into the report and the final report was then submitted to the Federal Joint Committee on 13 March 2006. This document, which has now been published on the Internet, has 314 pages and also contains the statements submitted and the minutes of the scientific hearing, in addition to the report itself. In the final report, IQWiG also discusses the results of the studies that were submitted together with the written statements but were not included in the evaluation.

Contact: Tel. +49 - 221/35685-0