Negative Pressure Wound Therapy ‚Äî A Descriptive Study
To address a persistent lack of evidence regarding the clinical outcomes of negative pressure wound therapy (NPWT) and identify which patient groups are most likely to benefit from NPWT, a retrospective, descriptive study was conducted to describe outcomes of this treatment modality when used in clinical practice.
Charts from a consecutive series of 87 patients (median age 68 years, range 16 – 92 years) who received NPWT during a period of 24 months were abstracted to a statistical software file. Patient demographics, history, and comorbidity variables as well as treatment outcomes were obtained from the computerized in- and outpatient record system. Treatment outcomes were grouped as successful (goal of care was met) or not successful (goal of care was not met). Successful treatment was noted for a total of 62 patients (71%) with a median treatment time of 17 days.
The proportion of patients with a successful outcome was significantly higher in patients with infectious, postoperative, and traumatic wounds than in patients with wounds related to peripheral vascular disease or pressure ulcers (P = 0.001). Treatment complications were observed in 18 patients (21%); five were related to infection. Quality-of-life concerns were noted as a reason for stopping treatment in four patients and equipment problems were recorded for two patients receiving NPWT in the home. This study confirms previous research that NPWT may be an effective and safe treatment method for acute wounds but further studies are needed to evaluate treatment efficacy and effectiveness in patients with peripheral vascular disease or pressure-induced wounds. Results also suggest that research protocols should include patient quality-of-life outcomes.
Potential Conflicts of Interest: none disclosed
Nonhealing wounds cause many patients to suffer. Treatment options have been limited to different kinds of wound dressings until the fairly recent development of new complementary methods, including negative pressure wound treatment (NPWT).
Only a handful of the numerous publications on NPWT are randomized controlled trials. Although several systematic reviews1-3 conclude there may be some evidence that NPWT is at least as good as treatment with modern wound dressings, the reliability and validity of these studies are questionable due to problems related to sample size, randomization methods, and allocation concealment. In addition, outcome measures and evaluation of the clinical importance of the different methods used are rife with discrepancy.
A few good quality studies on NPWT compare therapy outcomes in various wound etiologies. In a high-quality randomized clinical trial involving 60 patients with chronic leg ulcers, Vuerstaek et al4 showed a significant improvement (P= 0.0001) in time to complete healing using NPWT compared to standard wound care using hydrogel or alginate dressings. The type of wounds believed to benefit most from NPWT seem to be lower leg ulcers and split-thickness skin grafts.3 Llanos et al5 showed that use of NPWT following skin graft procedures significantly (P= 0.001) diminished the loss of split-thickness skin grafts and reduced length of hospital stay compared to control group with similar dressing but without connection to negative pressure. The authors of the latter study recommend routine use of NPWT following skin graft procedures. A recent consensus publication6 also favored NPWT for treating diabetic foot ulcers.
Due to a persistent lack of evidence, studies to evaluate the clinical outcomes of NPWT and to identify which patient groups are most likely to benefit from NPWT are needed. This is especially important because an approach that does not achieve wound treatment goals can increase suffering and decrease patient quality of life. The aims of this retrospective study are to describe basic demographic data, comorbidity, treatment results, and complications in relation to wound etiology in patients treated NPWT.
Material and Methods
Patients. All consecutive patients with wounds of any type of etiology treated with NPWT from 2005 to 2007 in a large general city hospital (Södersjukhuset, Stockholm, Sweden) were included in this retrospective study. The NPWT system used at that time was the vacuum-assisted closure (VAC) device (Kinetic Concepts, Inc., San Antonio, TX) using continuous subatmospheric pressure of 125 mm Hg. Dressings were changed two or three times a week or more frequently depending on the amount of fluid produced by the wound as judged by the clinician in charge (or when the canister was full). NPWT was used when conventional wound treatment failed or when no other options (such as for open abdomen) were available. Conventional treatments usually include advanced moist treatment with hydrofiber for wet wounds and nonadhesive dressings for dry wounds. Ethical approval for collection of data from patient files was obtained by the local Ethics Committee (2008/2023-31).
Baseline data. A protocol for collecting data from the patients’ charts was developed and data were abstracted by an experienced, specialized wound nurse who was not responsible for, or in charge of, any wound care patients. Basic demographic information and wound data were abstracted from the patients’ medical charts contained in the hospital’s computerized patient data recording system (Melior, Siemens AB, Upplands Väsby, Sweden) to an SPSS (Chicago, IL) data file. Uncertainties in chart data interpretation were discussed with a consultant general surgeon or a consultant orthopedic surgeon until consensus was reached. Consultations addressed the etiology classification of five cases. To evaluate interrater reliability, a nurse not involved in the study reviewed a sample (N = 9) of the medical charts and results were compared to those of participating clinicians. Information on the etiology of the wounds was grouped for further analysis as follows: 1) all types of open postoperative wounds after either orthopedic or general surgery (eg, open abdominal wounds, infections, and skin defects); 2) wounds related to peripheral vascular disease as confirmed by ankle brachial index or toe pressure measurement, including diabetic foot wounds and arterial leg ulcers; 3) wounds with an infectious origin such as necrotizing fasciitis and erysipelas; 4) trauma-related wounds that could not be treated with primary closure; and 5) pressure ulcers.
Information on smoking habits (yes/no), diabetes mellitus (yes/no), current cardiovascular diseases (yes/no), current alcohol abuse (yes/no), known peripheral vascular disease (yes/no), or ongoing dialysis treatment (yes/no) also was collected from the medical charts. Although smoking habits and alcohol abuse often were not reported in the medical charts, information on medical conditions could be regarded as reasonably reliable because this information is regularly noted by all physicians. If information about a medical condition was missing, the patient was classified as not having the disease or habit. All bacterial cultures taken during the healing process and the treatment with NPWT were documented.
Follow-up. All available follow-up information from the hospital in- and outpatient departments was abstracted and analyzed. The follow-up time ranged from 24 to 48 months. Treatment results were abstracted verbatim from the charts and grouped into the following categories: 1) successful treatment — eg, wound much improved and/or left to heal by secondary granulation; wound healed; wound bed improved and skin graft performed, and 2) nonsuccessful treatment — wound not improving, wound bed larger or worse, treatment discontinued due to complications. The duration of treatment (days) was calculated from the recorded treatment start and stop dates, and the reason for treatment discontinuation was noted as described. All forms of complications noted in the patients’ charts and related to the treatment were documented and categorized as 1) wound-related complication (wound condition worse, septicemia, bleeding, effect on surrounding skin in terms of maceration and skin irritation); 2) patient experience of the treatment (for example, tasks that had an impact on the patients quality of life, such as a decrease in food intake and problems sleeping due to NPWT); and 3) complication related to technical issues with the equipment.
Statistical analysis. Fisher’s exact tests were used to compare the demographic variables and the Mann-Whitney U-test was used to compare the treatment times of patients in the successful and nonsuccessful NPWT treatment outcome groups. Because the treatment time distribution was skewed, median times are presented.
Differences were considered significant if P<0.05, two-tailed. SPSS 17.0 (SPSS, Chicago, IL) was used for statistical analyses.
To calculate interrater reliability, all registered variables in the predefined protocol were compared between the chart reviewer and the external rater. All matched items were calculated as correct and in agreement. There was an 84% interrater reliability agreement between the chart reviewer and the external rater not involved in the study.
Patients. Charts from a consecutive series of 92 patients were abstracted. Results from three patients were excluded due to missing data and follow-up information was incomplete for two patients, leaving a sample size of 87 patients.
Basic patient demographic and wound-related data. Of the 87 patients treated, 53 (61%) were male and the median age was 68 years, range 16–92 years (see Table 1). The most common indication for NPWT use was the presence of a postoperative wound (45 patients — 22 orthopedic and 23 general surgical). Of the 23 general surgical wounds, 15 were open abdominal wounds. In 11 patients, wounds were due to peripheral vascular disease and included eight diabetic foot ulcers and three patients with lower extremity peripheral gangrene. Nine patients had wounds of infectious origin — six with necrotizing fasciitis and three with erysipelas; eight patients had traumatic wounds. Fourteen (14) patients had pressure ulcers; 11 had an ulcer in the pelvic region and three had ulcers on the foot.
Sixty-three patients (72%) had a history of an underlying medical condition or a comorbidity; the most common comorbidity was cardiovascular disease (31 patients, 36%), followed by diabetes (26 patients, 30%). Patients with wounds due to peripheral vascular disease had the highest frequency of comorbidity (100%).
Bacterial swabs were taken at the beginning of treatment and during treatment if secondary contamination or infection was suspected; as many as 81 wounds (93%) had a positive bacterial culture during treatment with NPWT. The most common bacterial species was Staphylococcus aureus (31, 36%) (see Table 3).
Treatment outcome.Sixty-two patients were classified as having a successful and 25 as having a nonsuccessful treatment outcome (see Table 4). Of the unsuccessfully treated patients, nine showed no signs of reduction in wound size or any positive effect on wound healing and 16 patients suffered complications during NPWT. There were no significant differences between the successful and nonsuccessful treatment groups regarding gender, age, comorbidity, or bacterial culture results.
Wound etiology had a significant impact on the treatment outcome (P= 0.001). The highest rate of successful treatment (100%) was seen in patients with infectious wounds, followed by postoperative wounds (37, 82%). All postoperative wounds with a treatment goal of secondary closure and where NPWT was used as preparation for skin graft had a successful outcome (P= 0.01). The lowest rate of successful treatment occurred in patients with peripheral vascular disease, where only three wounds (27%) were recorded as having a successful outcome.
No significant differences were noted in the length of treatment between wounds with a successful (median 17 days, range 2–210) and an unsuccessful outcome (median 19 days, range 2–75). However, treatment time and treatment results differed between wounds with varying etiology (see Table 5). A significant difference between treatment time and treatment results was noted in wounds caused by trauma — wounds with successful outcomes required 6 days of treatment compared to 27 days for wounds treated without success (P= 0.04).
No significant differences were noted between the postoperative wound patients with pressure wounds and patients with wounds due to peripheral vascular disease. The patients with pressure ulcers had the largest range for treatment time, with a span of 8 to 210 days of treatment.
Complications related to the treatment were noted in a total of 18 patients (21%); among these, the outcome was successful in two (11%) and unsuccessful in 16 (89%); P = 0.001. Wound complications were documented in 10 patients — five developed an infection, three had deterioration of the surrounding skin, and two developed bleeding/hematoma. Treatment was discontinued at the request of four patients and reported as discontinuation of NPWT related to a severe deterioration in quality of life and/or negative impact on their daily life. In one patient, treatment was discontinued due to a decrease in general health status, particularly cognitive status, which made NPWT treatment problematic. In one patient, the physician terminated the treatment due to suspicion of development of a bowel fistula; in two other patients, technical problems with the equipment and difficulties handling the device at home resulted in discontinuation. Out of these 18 patients, treatment was terminated in 14 patients before achieving treatments goals.
The results of this study show successful wound outcomes — eg, closure and/or decrease in wound size — were achieved in the majority of patients treated with NPWT (71%). The highest rate of successful treatment outcomes was seen in wounds caused by infection (median 11 days), followed by postoperative wounds (82% of patients, median treatment time 21 days). The lowest rate of successful treatment (27%) was observed in wounds caused by peripheral vascular disease.
One of the difficulties with comparisons between studies is that the primary outcome and follow-up times vary considerably. In this study, the outcome was categorized as successful or not successful based on the goal of treatment because in clinical practice, NPWT is used to help achieve a variety of wound care goals, including preparing the wound bed for an intervention such as skin transplantation, improving wound status, or to palliate and alleviate symptoms. Although other outcome measures such as reduction in wound size provide important efficacy information in controlled studies, they do not reflect a patient-oriented outcome of care.
The results of this study confirm the retrospective study results from Clare et al,7 who indicated that NPWT may not be suitable for patients with severe peripheral vascular disease and small diabetic foot ulcers. A review of a randomized controlled study3 that reported good outcomes in patients with diabetic foot ulcers included only patients with adequate peripheral circulation.
NPWT used for pressure ulcers also resulted in less successful outcomes in this study. This finding is similar to a results of a prospective RCT8 and a review of an RCT,9 where no evidence of NPWT effectiveness for treating pressure ulcers was observed. In the current study, patients with pressure ulcers and a successful result required an extremely long treatment time (median time 28 days, maximum 210 days). This finding also has been documented in a meta-analysis of published data.10
Adverse events and complications when using NPWT often are reported as rare or nonexistent. However, in this study, treatment was discontinued in 20% of patients.
The most commonly reported complication in a review of an RCT3 is wound infection; this also was the case in the current study where five patients (6%) developed a wound infection. In addition, one third of treatment terminations were requested by patients and related to a reported decrease in their quality of life as a result of the treatment. This finding suggests the impact of NPWT on a patient’s daily life is probably quite substantial. Because existing research has not focused on this issue, studies regarding knowledge about patient perception of this treatment and quality of life are needed.
The observed outcomes from this retrospective study also suggest that successful outcomes within a reasonable period of time are etiology-dependent. Wounds caused by trauma appear to respond rapidly, whereas the potential advantages of using NPWT in other wounds seem less pronounced. The observations in wounds secondary to pressure and peripheral vascular disease in particular confirm the need for a careful reevaluation of NPWT outcomes after 2 weeks of treatment11 and that wound etiology should be taken into account when deciding on subsequent treatment measures.
Further research is necessary to evaluate the efficacy of NPWT in achieving the goals of care for different types of wounds, especially for wounds related to peripheral vascular disease, because results from previous RCT reviews3,6 and a retrospective study7 are contradictory.
Study Strengths and Limitations
One of the strengths of this retrospective study is that data are based on a consecutive series of patients; all patients treated in a given period with wounds of varying etiology were included. These inclusion criteria can be generalized to provide a broad and representative picture of patients treated with NPWT in hospitals.
One weakness of the study is that data are based on information retrieved from patients’ charts. However, even though this method has important limitations, validity can be improved, as shown by Panacek,12 who noted that by following “the ten commandments” of chart review research, validity can be enhanced and possible biases limited. This study design included nine of the 10 recommended steps; only blinding was not achieved.
To ensure study reliability, an interrater reliability test was performed. Gilbert13 offers ways to enhance reliability by using a second reviewer to re-abstract a sample of charts, blinded to the information obtained by the first correlation reviewer. In this study, a nurse not involved in the study reviewed a sample of the charts and achieved an 84% agreement between raters, an adequate interrater reliability score.
In this study, comorbidity information that was missing in the medical charts was interpreted as if the patient did not have the condition; this could result in underestimating the impact of these factors on the treatment result, especially in subgroups with a small sample size. The results from these small subgroups provide important information but should be interpreted with caution.
Finally, conclusions about the potential effectiveness of NPWT in this population are limited by the absence of a control group.
This study confirms previous research suggesting that NPWT may be an effective and safe treatment method for several types of wounds, especially infectious, traumatic, and postoperative wounds. However, because this study did not have a control group and included very few patients in the different subgroups, the results should be interpreted with care. Controlled studies focusing on different patient groups are needed, particularly because there seems to be a disagreement concerning the success of NPWT for patients with wounds due to peripheral vascular disease. Current findings also suggest that future studies should include patient quality-of-life variables.
Ms. Wallin is an OR-nurse and Dr. Boström is a general surgeon, Department of Surgery, Department of Clinical Science and Education, Södersjukhuset, Karolinska Institutet, Stockholm, Sweden. Dr. Ulfvarson is an RN, Department of Neurobiology and Society; and Dr. Ottosson is an orthopaedic surgeon, Department of Orthopedic Surgery, Department of Clinical Science and Education, Södersjukhuset, Karolinska Institutet. Please address correspondence to: Ann-Mari Wallin, RN, Department of Clinical Science and Education, Södersjukhuset, Karolinska Institutet Södersjukhuset, S-118 83 Stockholm, Sweden; email: firstname.lastname@example.org.