A comparative analysis of pressure sore treatment modalities in community settings

The management of pressure sores in community settings, poses a clinical problem which challenges the patient’s toler­ ance and the clinician’s diligence and ingenuity. Pressure sores can be painful, lead to infection and are associated with con­ siderable morbidity and increased mortality (Patterson & Bennett, 1995:919; Bale, Banks, Hagelstein & Harding, 1998:65). Treatment costs of these wounds are high in terms of resources (Colin 1995:65; Wood, Griffiths & Stoner, 1997:256). However, since there are untold cost in terms of pain and suffering to the patient, it is impossible to calculate the true cost of pressure sores (Dealey, 1994:87).


Introduction
The m anagem ent of pressure sores in community settings, poses a clinical problem which challenges the patient's toler ance and the clinician's diligence and ingenuity.Pressure sores can be painful, lead to infection and are associated with con siderable m orbidity and increased m ortality (Patterson & Bennett, 1995:919;Bale, Banks, Hagelstein & Harding, 1998:65).Treatment costs of these wounds are high in terms of resources (Colin 1995:65;Wood, Griffiths & Stoner, 1997:256).However, since there are untold cost in terms of pain and suffering to the patient, it is impossible to calculate the true cost of pressure sores (Dealey, 1994:87).
Patients in a community setting are as vulnerable to sustaining pressure sores as those in hospital (James, 1997:12).Several community risk factors are likely to present problems to skin integrity, namely immobility and unrelieved pressure, malnutri tion, age, incontinence and chronic illness.Due to the increase in the ageing population and the associated higher degree of morbidity, along with the em phasis on provision of care shift ing towards the earlier discharge o f acutely ill patients, it is expected that the number o f patients at risk for developing pressure sores in the community will increase (Inman & Firth, 1998:515;Glover, 2000:161).This phenomenon will add to the at-risk-population in the community, thereby placing a greater demand on community resources as a result of an increase in the incidence of pressure sores.Consequently it is clear that the most cost-effective and acceptable method of wound m an agement is needed to address the problem of large numbers of pressure sores in the community.
The most important aspects o f managing patients with pres sure sores include relieving pressure and shear force or fric tion, promoting continence and maximizing nutrition and fluid intake.However, local wound m anagement does have an im portant role to play.A cost-effective method of wound man agement implies a method that will result in rapid wound heal ing.Since W inter's landmark research (1962:294) it has been recognized that a m oist w ound environm ent prom otes reepithelialisation and healing, thereby supporting cost-effec tive wound care management.Despite this evidence many cli nicians still use alternative treatment methods that cause des iccation o f the wound-bed and thereby impede the healing process.
Since wound healing is an intricate and dynamic process, it has been suggested that no single dressing is suitable for the management of all wounds, particularly at all stages of the healing process (Bux & Malhi, 1996:305).However, the major ity of modern dressings are now designed to maintain a moist environment at the wound interface, providing conditions for rapid epithelialisation and thereby improved healing (Frantz & Curationis Gardner, 1994:39;Bux & Malhi, 1996:305;Dale, 1997:12).Though many o f these m odem dressings are more expensive than tra ditional cellulose-based products, very often wounds dressed with newer products, heal more rapidly than those dressed with conventional materials resulting in cost savings (Tho mas, 1997).
The choice of dressing or treatment method may be influenced by several factors -a major one being the cost.Another influ ence on the choice of dressing or treatm ent method is its ac ceptability to patients and caregivers.Often patients with chronic wounds and their caregivers have firm views on the dressings that they will accept based upon prior experience with these or similar materials (Thomas, 1997).
Controlled clinical wound care trials are an important part of evaluating the performance o f wound dressing materials and provide research-based information to nurses and clinicians striving to establish effective treatment strategies (Banks & Bale, 1994:301).

Purpose
The purpose of this research was to compare current wound care managem ent methods with a more advanced wound care method using three Smith & Nephew™ wound care products, in the treatm ent o f patients with pressure sores in the com m u nity over a six week period.

Objectives
Specific objectives included comparison o f cost-effectiveness with regard to treatment cost and rate o f healing as well as assessment o f the acceptability of these treatment modalities to patients and caregivers in terms o f ease of application, com fort o f the dressing, durability and ease o f removal.Inform a tion regarding wound sites and patient characteristics was also collected.

Research design
O f the nine main experimental designs available to the nurse researcher, as identified by Bums and Grove (1993:317-323), a prospective randomized, controlled trial was best suited to phenomenon under investigation.This experimental design may further be classified as an open-label clinical trial since both the researcher and the patients knew the nature of the treatment.In support of the chosen design the following was noted.E x perimental designs provide a significant amount of control in order to examine causality more closely.It has also been pro posed that this technique be used to examine areas o f nursing practice, such as com paring traditional nursing care practices with newer techniques (Bums & Grove, 1993:323).

Population Patient recruitm ent
Patients were recruited via referrals from primary health care clinics, comm unity health care workers and other health care professionals practicing in the community.Several patient re cruitment strategies were used namely: • Communicating with colleagues directly, via E-mail and telephone requesting referrals.
• Speaking at formal and informal professional meet ings requesting referrals.
• Placing notices in places where colleagues and/or pa tients would see them.

•
Placing advertisem ents in local newspapers.Over a tw elve-m onth period (from May 1999 to June 2000), a total of 60 patients were screened for inclusion to the study, of whom 58 patients met the inclusion criteria and were consid ered suitable for entry.

Inclusion criteria
• Patients in the Bloemfontein community aged 18 years or older with a clinically uninfected Stage 2 ,3 or 4 pressure sore.(The Stirling Pressure Sore Severity Scale (Waterlow, 1996:56) was used as classification system ) • Patients or their guardians, who gave informed con sent.
• Patients who were willing and able to comply with the treatment.

Random isation
Eligible patients were random ised into an equally sized con trol-and experimental group.Randomisation was done by pres sure so re S ta g e , a c c o rd in g to a c o m p u te r g e n e ra te d randomisation list provided by the D epartm ent Biostatistics, University o f the Free State.If a patient had more than one pressure sore, all sores were treated with the same treatment modality.However, only one sore was chosen at random for inclusion in the study and it was on the basis of this sore's classification, that random isation occurred.In order to select a sore at random, the sores were numbered from top (superior) to bottom (inferior), left to right and a random num ber selected from a random num ber table.
Unlike institution-based studies, com m unity-based studies do not allow for larger sample sizes without certain implications (Bale & Banks. 1994:304).These implications refer to practical issues, such as variable control and the financial implications of extensive com m unity-based studies that require additional skilled field workers.In view of these implications and the docu mented successes o f smaller com parable studies by Bale andco-workers (1997:463 and1998:65), it was decided in collabora tion with biostatisticians that a sample of at least 40 patients 'w as a statistically adequate number.

Treatment modality
Two treatm ent modalities were used namely: an advanced wound care management method (using Smith & Nephew™ products) for patients in the experimental group and the cur rently used management method for patients in the control group.

A d va n c e d wound care m anagem ent
The peri-wound area was cleansed with a gentle, hypoallergenic soap and water to remove any excessive exudate and transient m icro-organism s from the surrounding skin and dried with gauze.The wounds were than aseptically cleansed with warm (approxim ately 37°C) sterile, physiological saline.A gentle cleansing technique was used by either irrigation, if the wound had delicate granulation tissue, or by gently patting the wound bed.
The three trial dressings were the following: A hydrogel (IntraSite™ gel) A foam dressing (Allevyn™ hydrocellular polyurethane sheets or Allevyn™ adhesive) A transparent film dressing (OpSite Flexigrid™) W here non-viable tissue was present on the surface of the wound requiring debridement, a thin layer of IntraSite™ gel w as applied on the w ound bed and covered w ith either Allevyn™ non-adhesive, hydrocellular sheets or Allevyn™ adhesive.If the wound was granulating and no non-viable tissue was present at the surface, then Allevyn™ non-adhe sive hydrocellular sheets or Allevyn™ adhesive were applied directly to the wound.Transparent OpSite Flexigrid™ dress ings were applied to Stage 2 epithelialising wounds.OpSite Flexigrid™ was also used as a secondary dressing to secure Allevyn™ non-adhesive dressings.All the selected dressings were applied to the wounds strictly according to the m anufac turer's instructions.

C u rre ntly used wound care m anagem ent
The peri-wound area was cleansed in the same way as those in the experim ental group.The wound was then aseptically cleansed and covered with the available wound care materials and/or methods as used on that particular patient at that time by the patient or primary caregiver.These materials included cleansers such as saline and water; a variety of disinfectants such as hypochlorite solutions (Milton, Eusol), hydrogen peroxide, cetrimide, chlorhexidine; antibac terial ointments and creams such as povidone iodine and silver sulfadiazine; systemic antibiotics such as metrinidazol that were crushed and sprinkled over the wound (bear in mind these were non-infected wounds); essential oils such as Tea Tree oil and St John's wort; pure honey; mixtures/potions that amongst others included a mixture of povidone-iodine, honey and chlo ram phenicol.D ressings included cotton w ool, alginates, hyrocolloids, gauze impregnated dressings (m edicated and non-medicated) and gauze.
In order to avoid cross-contamination and subsequent wound infection, the researcher maintained strict adherence, in both treatm ent modalities throughout the entire study, to the basic principles of asepsis and infection control, as recom m ended by the Agency for Health care Policy and Research (AHCPR) (1994:1-10).Similarly, throughout the study, the researcher pro vided continuous education and encouraged all patients and caregivers to adhere to the measures for pressure reduction, daily skin examination and interm ittent pressure relief tech niques as described in the Clinical Practice Guidelines by the AHCPR (1994:1-10).Additionally the researcher attempted, where possible, to address those identified factors that could delay wound healing.The frequency of dressing changes was determined by the unique nature and needs of each patient's wound.

Period of trea tm ent
Each patient was managed for a period of six weeks or until one of the following end points were reached: • the wound healed; • the patient withdrew; • an adverse event occurred in which the treatment benefit was unacceptably inferior to treatment risk.
four domain experts for comment regarding appropri -ateness, accuracy and representativeness of items on the forms.Once these comments were returned, ap propriate changes to refine the forms were made thereby assuring content-related validity evidence.
Mono-operation bias to assess wound healing was ad dressed by the use of multiple evaluation methods thereby improving the construct validity.
Reliability of the instrument was improved through consistent use of it throughout the data collection proc ess.This was facilitated by the use of only one data collector thereby avoiding bias and inconsistency and problems associated with interrater reliability.
Random assignment to experimental and control groups was used as a design strategy to control extra neous variables and lack of equivalence.

Data collection
Wounds were assessed weekly and data collected using spe cifically designed data collection forms.Additionally four evaluation m ethods were used to assess the rate of wound healing namely: • standardized digital wound photography; • tracing of the wound edges; • measurements of the wound and its appearance as well as • descriptive field notes.

Data analysis
The data analysis included a cost analysis and statistical analy sis.Cost-effectiveness was expressed by dividing the number of patients that healed into the total cost of treating that par ticular group.This indicated the cost of healing per patient healed.
Demographic and baseline inform ation was summarized by group.Numeric variables were summarized by medians and percentiles as the distributions were skew.Categorical vari ables w ere sum m arized by frequencies and percentages.
Changes betw een baseline and consecutive treatment week information were summarized per group by medians and per centiles or percentages, as appropriate for the difference be tween the groups, with 95% confidence intervals.The percent age of withdrawals and adverse events were compared be tween the two groups using 95% confidence intervals for dif ferences in percentages.The log-rank-sun'ival test was used to calculate the percentage of patients that healed by the end of each week.

Validity and reliability
To increase the validity and reliability of the study, several approaches and techniques were applied.They included the following: • Preliminary data collection forms were forwarded to Patients were homogenous in terms of three extrane ous variables namely the presence o f an uninfected pressure sore, aged 18 years or older and living in the Bloemfontein community.
A pilot study was undertaken which provided the re searcher with the opportunity to test and improve the data collection forms and become familiar with the photographic equipment and technique.
The study and treatment methods were continuously monitored by an official, independent study monitor thereby improving the validity and reliability of the study.

Ethical considerations
The main ethical considerations referred to the right to self determination, confidentiality and protection from harm.
Patients were asked to partake voluntarily and were only en rolled into the study with written informed consent, provided by themselves or their legal guardians, and where possible in collaboration with their attending physician(s).All participants were asked to sign a consent form, which contained a detailed description of the trial procedures and methods.Consent forms were available in the three main languages spoken within the Bloemfontein community namely Afrikaans, English and South-Sotho.In addition patients had the right to withdraw from the study at any time without penalty.
All patient names were kept confidential.The study num ber allocated to them during the study identified patients through out the evaluation and documentation.Patients were carefully monitored and where any adverse event occurred, the patients were discontinued and managed appropriately.
The study was performed in accordance with the guidelines of the World M edical Association Declaration of Helsinki (1996) and continuously monitored by an official study monitor.Af ter obtaining approval from the Ethics Committee, Faculty of Health Sciences, University o f the Free State, the study was initiated.

Results
Sixty patients were screened for inclusion in the study, of whom 58 p atien ts (N = 58) w ere co n sid ered su itab le for entry.The majority o f pressure sores in both the experimental group (70%) and control group (60,4%) were located on the sacrum or trochanters.However, the site with the highest frequency of ulceration was the sacrum.There were no significant differ ences between the two groups on entry therefore the two groups were equivalent in terms o f dem ographic and prognostic fac tors and thus comparable.
dents, defined as any undesirable clinical occurrence in a pa tie n t w hich w as th o u g h t to be d ire c tly re la te d to the investigational product(s), occurred.

Healers and non-healers
The total number of patients who completed the study, that is, healers and non-healers, was 41.This represents 70,6% o f the total study population (N=58).

W ithdraw als
Fifty-eight patients (N=58) were enrolled in the study in week zero.However, during the course of the six-week treatment period a total o f 17 (29,3% ) patients were withdrawn as they died, moved from the geographical area, developed a wound infection or were hospitalized.(See Table 2) No adverse inci- The wounds of a significant number of patients 15 (53,6%) in the experimental group were healed as opposed to only 9 (30%) in the control group.Notably all the wounds that healed in both groups were classified upon entry as stage two pressure sores.The log-rank-survival test was used to offset the higher death rate in the control group and make m axim um use o f all available patient data up to the point o f withdrawal.Results of the log rank test, depicted in Table 3, indicate the percentage of patients healed by the end of each week.Although there ap peared to be a faster rate of healing in the experimental group than in the control group, the difference was not statistically significant (p = 0.15).An alm ost equal num ber of patients in both groups [8 (28,6%) in the experimental group and 9 (30%)] remained for the entire study duration but did not achieve heal ing.
The total cost of treating patients in the control group who completed the study was R22 824.71.Nine out of 18 patients in this group were com pletely healed.W hen dividing the healed group (9) into the total treatment cost, the cost to achieve healing in the control group was R2 536.08 per patient.Sim i larly the 95% confidence interval, [1644.4; 3427.7],here indi cates the range o f treatm ent co st f o r the control group only.Since the values of the confidence intervals for the re-

Com parison of treatm ent cost
Despite the fact that there were no significant differences be tween the two groups as far as cost is concerned, it is evident from the 95% confidence interval [-789.75; 159.12] that there was a tendency for the control group to have higher values indicating that the control treatm ents tended to be more expen sive than the experimental treatment.

Cost-effectiveness
The total cost of treating patients in the experimental group who completed the study was R23 052.02.spective groups overlap, there was no statistically significant difference between the groups as far as cost-effectiveness was concerned.However, there does appear to be a tendency for the control treatm ent to be more expensive and thus less costeffective than the experimental treatment.

Dressing acceptability
The acceptability of the wound treatment method to patients and caregivers was assessed on completion of the study pe riod by means of a structured questionnaire.

P a tie n ts ' assessm ent of dressing acceptability
Patients indicated their responses to specific statements re lated to the comfort of the dressings on application and re moval (See Table 4).Fourteen (60,9%) of patients in the experi mental group agreed that application was comfortable as op posed to only 6 (33,4% ) patients in the control group.No one in the experim ental group experienced discom fort on dressing removal.However, one (5,6%) patient in the control group ex pressed strong discom fort on dressing removal.Notably a large number of patients [9 (39,1 %) in the experimental group and 11 (61,1%) in the control group] could not report on the comfort of the dressing, as they were unresponsive due to age-related dementia.

Conclusion
From the results it may be concluded that over a six-week pe riod more wounds in the experimental group, treated with ad vanced wound care products and in particular three Smith & Nephew ™ products, healed, than wounds in the control group

C aregiver' s assessm ent of dressing acceptability
The overall ease o f dressing application in the experimental group was m uch higher.This was shown to be statistically significant: 95% confidence interval [1,6; 47,7].Although the caregiver did not find it difficult to remove any dressings, it does appear that dressing removal in the experim ental group was easier than in the control group (See Table 5).

C aregiver' s assessm ent of dressing durability
Dressing durability over the six-week treatment period was assessed by m eans o f a rating scale.The scale ranged form one (not durable) to five (extremely durable).The assessment showed that dressings used to treat patients in the experimen-treated with the currently available wound care products.
As there were no significant statistical differences between the groups with regard to the cost to achieve healing, it may be assumed that the advanced wound care treatment method was not more expensive than the currently used methods.How ever, there was a tendency for the currently used methods to be more expensive.Furthermore, when cost of treatment for all patients who completed the study (healers and non-healers) was compared results indicated a tendency for the advanced treatm ent method, using the three Smith & N ephew ™ prod ucts, to be less expensive.Consequently it may be concluded that the results of this study indicate a trend for advanced wound care methods to be more cost-effective than currently used methods.A lthough these trends were not statistically significant, they could be of clinical and practical importance.
With regard to acceptability of dressings in terms of ease as well as comfort of application and removal, the majority of pa tients in the experimental group rated their dressings as ac ceptable.In contrast only a third of patients in the control group rated their dressings as acceptable.Finally, the caregiv e r's assessment of dressing acceptability in terms of ease of application and removal as well as durability over the six-week period indicate that dressings used in the advanced wound care management method (Smith & Nephew™ products) were significantly more acceptable than dressings used in the cur rent management methods.

Recommendations
It appears that the advanced treatment method lends itself to better wound care practice with potentially financial benefits to individuals and institutions, yet there is still a tendency for currently used traditional treatm ent methods to persist.Con sequently a paradigm shift towards a more advanced wound care treatment method is needed.However, this will require training and education about wound care to patients, caregivers and all health care professionals.
During the 12-month data collection period of this study the researcher saw evidence of an alarming lack of knowledge with regard to all aspects related to the basic principles of pressure sore prevention and treatm ent, not only in patients and caregivers but also health care workers which included profes sional nurses and physicians.The development of a com pre hensive best practice model for pressure sore prevention and treatment in community settings could be of an enormous value to wound care practitioners and significantly benefit patients with pressure sores in the community.
Despite advances in wound care products and much interna tional research on the topic over the past decade, a literature search revealed no published community-based wound care trials in South Africa.As such this study accentuated the need for more clinical research in wound care in South Africa.

Table 1 :
Base-line demographic dataRandomisation resulted in 28 patients allocated to the experi m ental group and 30 patients to the control group.Base-line dem ographic data (See Table1) acquired on entry were: gen der, age, allergies, weight, height (body mass index) and wound site.Assessment data acquired on entry and thereafter at weekly intervals included: pressure sore risk assessm ent (using the Braden Risk Assessment Scale[Bergstrom, Demuth & Braden,  1987:417]), wound duration since onset, wound dimensions, pressure sore stage, level of exudate, appearance o f the wound

Table 2 :
Population (N = 5 8 ) status at the end of the study period

Table 3 :
Percentage healed by the end of each week

Table 4 :
Patients'assessment of dressing application and removal

Table 5 :
Caregiver's assessment of dressing application and removal