Platelet-rich Plasma in Skin Ulcer Treatment

Raquel Cobos Campos, Pharm; Naiara Parraza Diez, Pharm, PhD; Felipe Aizpuru Barandiaran, MD, PhD


Wounds. 2013;25(9):256-262. 

In This Article


A cost-effectiveness analysis comparing 2 treatment options—usual clinical practice and PRP—for the treatment of skin ulcers after 48 weeks was done. Markov-based modeling that allows simulations of complex systems,[17] such as health care processes, was used. The model was constructed following a previously established protocol, using estimations obtained from published or other available data on efficacy, toxicity, and costs of the various options being compared.[17]

The clinical course of the ulcers was modeled using a hypothetical cohort of 200 patients with characteristics similar to those included in the clinical trials comparing the 2 options reviewed in this study (conventional treatment and PRP): patients with diabetes between the ages of 55 and 75 years, who have at least 1 ulcer on the lower extremities that had been detected at least 4 weeks earlier, and with a current area of 3 cm2 - 10 cm2. These 200 patients were randomly allocated to receive either 1) usual care: cleansing of the wound; debridement of necrotic tissue; prevention; diagnosis; and, as required, treatment of infection and the application of suitable dressings; or 2) PRP treatment: cleansing of the wound; debridement of necrotic tissue; prevention; diagnosis; and as required, treatment of infection and application of PRP as well as suitable dressings.

The treatment applied to these hypothetical cohorts consisted of 12-week cycles with 2 appointments per week. At each appointment, the wound was cleaned appropriately and covered with a suitable dressing. Additionally, the intervention group received PRP on 1 of the 2 days, always on the same day of the week. Once the first 12-week cycle of treatment had been completed, patients in whom complete healing had been achieved were considered to have finished their treatment and were monitored for the remaining 36 weeks to detect potential recurrences. Patients whose wound had not completely healed after 12 weeks, in either of the groups, continued with the conventional treatment until the ulcer was fully healed. Similarly, patients in whom the ulcer recurred after healing received the conventional treatment until the new ulcer was fully healed.

Estimation of Effectiveness

To obtain data on effectiveness, a literature search was conducted in secondary databases PubMed, EMBASE, and Cochrane, using the following strategy: ("wound healing[mesh]" OR "leg ulcers[mesh]" OR "skin ulcers[mesh]" OR "venous ulcers[mesh]") AND ("plasma rich in growth factors[mesh]" OR "platelet rich plasma[mesh]") AND ("clinical trial[mesh]" OR "controlled randomized trial"[mesh]).

Only published original papers and unpublished preliminary results provided by researchers describing studies involving a control group that received conventional treatment (ie, wound cleansing, debridement, infection management, and application of suitable dressings) and lasting for 8 to 12 weeks were selected. No restrictions were applied in terms of language or date. Research that considered a measure of effectiveness other than the number of healed ulcers at the end of the study were excluded.

A meta-analysis was performed on the data in the selected papers to obtain combined effectiveness results after 12 weeks for each of the treatment options. This was used as a starting point to estimate the probability of healing at 48 weeks.

For this estimate, the authors used a Markov model with 4 possible levels of health status: nonhealed ulcer, healed ulcer, recurrence (of the ulcer), and amputation (Table 1 and Table 2), and 4 cycles of 12 weeks each. The probability of healing was different in each cycle, while the period for which each patient had a given health status depended on their treatment group allocation and the time elapsed since the start of the treatment.[18]

Each patient in each treatment group began the study with the status of "nonhealed ulcer." During each cycle in which the patients had this status, their ulcer might heal or not. In the latter case, amputation might then be required. If patients had the status of "healed ulcer," they either continued to have no ulcers or their ulcer recurred. Table 2 summarizes the probability of changing between health status categories.

Estimation of Costs

The economic analysis was performed from a health system perspective. The authors only took into account direct costs based on nursing staff wages from Lazaro-Martinez et al[19] for the Spanish publish health care system in 2007, and converted into euros; the mean cost of a PRP kit per treatment calculated from the price of 4 different formulations, in some cases from quotes provided by the corresponding supplier, and in others, as advertised on the company website; costs of the materials necessary for changing dressings, taken from Ragnarson and Hjelmgren,[20] using the average between costs in Sweden and the United Kingdom; and costs of amputations[21] due to ulcers not healing. Table 3 shows a breakdown of all costs included in study.

Statistical Analysis

The heterogeneity of the studies selected was examined with the DerSimonian and Laird method,[22] assessing the level of variability in the results across the studies. Given the low power of the test, the authors decided to reduce the level of confidence to 90%.

The authors calculated the number needed to treat (NNT) at the 95% level of confidence (the number of ulcers to be treated using PRP to achieve a higher rate of healing than that with usual care), and the incremental cost (the increase in costs associated with adding PRP to the usual care to heal 1 more ulcer). The measure of effectiveness used was the number of ulcers healed after 48 weeks, expressed in absolute numbers, with the percentage distribution and 95% confidence intervals (CIs), and the costs were calculated in euros. It was not considered necessary to adjust costs or benefits to present values given that the study period was less than 1 year.[20]

A sensitivity analysis was performed recalculating the NNT and incremental cost with the CI percentages corresponding to the measure of effectiveness for each treatment option for which the difference between the outcomes with the 2 treatments was the smallest.