Articles » Injuries » Other » Guidelines for PRICE » 8 - Evidence

8 - Evidence for the guidelines

Here follows a summary of the guidelines for the management of soft tissue injury during the first 72 hours, together with an evaluation of the supporting evidence. Each section will conclude with a brief summary of current practice with respect to each of the elements of PRICE, as extracted from a questionnaire distributed to 500 members of the ACPSM. A summary of the questionnaire response can be seen in Appendix E.

The guidelines are for the application of PRICE; they are based on the assumption that an accurate diagnosis has been made, and that PRICE is the appropriate intervention of choice.

Individuals may decide to apply alternative modalities, as identified in Appendix C. Individual consultation and agreement may be required to ensure adherence to the guidelines.

Situations in which caution should be observed have been identified in sections 1.2, 3.4 and 7.7.1. Specific contraindications are identified in individual statements relating to each of the elements of PRICE. A summary of precautions and contraindications are presented in Appendix G.

8.1 - Protection (and prevention of further injury)

Injury results in a sudden drop in the tissue’s ability to withstand tensile stress. Immediately following injury there is little or no drop in tensile strength but within the first few days, as the inflammatory process evolves, a significant loss of tensile strength is observed (Houglum, 1992). The extent of this loss is proportional to the degree of tissue damage.

Protection against further tension to the area of tissue damage is essential until an accurate diagnosis of the extent of the lesion is established (Hunter, 1994). In the early days following injury (up to four to six days), tension to the site of injury may disrupt the fragile fibrin bond which forms a network of ’scaffolding’ which provides the link between the margins of the injured tissue (Hunter, 1994).

Protection may vary from the initial moving of the injured athlete from the location in which further injury may occur, through provision for general support to avoid using the injured part (eg crutches), to the application of specific means of limiting movement (eg braces / splints).

Various modes of protection have been advocated:

• Crutches (Pincivero et al, 1993; Reider et al, 1993)
• Slings (Levin, 1993)
• Taping (Pincivero et al, 1993)
• Hinged brace (Reider et al, 1993; Cross et al, 1990)
• Plaster of Paris / splints

Protection does not form the focus of any of the papers reviewed. However, it is mentioned as an element in the total early management of soft tissue injuries (including following surgery) but specific recommendations are not provided.

However, taking into consideration the nature and timing of the healing process and the evidence of the effect of excessive early stress on healing tissue, the following recommendations are made.

Guideline 1 - Protection

1 - Protection should be applied during the early stages of the healing process (at least up to day three) (II).

2 - The duration of application should be dictated by the severity of pain and the extent of injury. Animal studies suggest that a moderate (second degree) injury requires three to five days protection. Mild (first degree) injuries may require a shorter period and severe (third degree) longer (II).

3 - The mode of application of protection will depend on the site and nature of injury. This may range from protection from full weight-bearing (crutches), to general support (slings), to specific support for the injured structure/s (braces, splints, taping) (C).

4 - Whilst supporting/ protecting the injured structure/s, the mode of protection should avoid complete immobilisation of the part whenever possible (C).

5 - The mode of protection must be capable of accommodating oedema (C).

Evidence

Animal studies provide a sound biological base to support the use of some form of protection to avoid excessive stress on healing tissue during the first three to five days of the healing process (Jarvinen and Lehto, 1993; Burrough and Dahners, 1990; Buckwalter, 1995).

Although several clinical studies have investigated the concept of rest/ immobilisation versus exercise/ mobilisation following soft tissue injury, no reports of clinical trials investigating the efficacy of different modes and duration of application of protection were found. Consequently, empirical evidence on the efficacy of protection is lacking.

In a RCT to compare the effectiveness of early functional and conventional treatment for acute ankle ligament injuries, the use of some form of protection in both groups was advocated by Karlsson et al (1996). The “functional” group wore compression pads with elastic bandage followed by ankle tape, while the control (”conventional”) group had elastic bandage with crutches. The “functional” group had shorter sick leave and earlier return to sports, but there was no difference in the final outcomes between the two approaches, both of which demonstrated good results.

Recommendations for guidelines

There is general acceptance of the use of some form of protection during the early stages following soft tissue injury (Paessler et al, 1992; Pincivero et al, 1993; Levin, 1993; Reider et al, 1993; Karlsson et al, 1996); evidence for its specific use tends to be by consensus.

Current practice - results of questionnaire

When asked if they would prescribe protection to a second degree soft tissue injury during the first 72 hours following injury, 96 per cent of those who returned the questionnaire responded to this question. Of these, 94 per cent stated that they would prescribe protection whilst six per cent stated they would not. Crutches and slings were the most common mode of protection prescribed with 89 per cent (82 per cent of the total returned) stating they usually or sometimes prescribed crutches and 70 per cent (56 per cent of total) stating they usually or sometimes used slings. Inflatable and non-inflatable splints were less commonly prescribed, with only 17 per cent of the total prescribing inflatable and 38 per cent prescribing non-inflatable splints. This may reflect the availability of the splints and their cost.

8.2 - Rest

The appropriate role of activity in the management of injuries has been a subject of considerable controversy. The long-accepted concept that physical activity through loading and movement alters the structure of the musculoskeletal structures has prompted supporters of early post-injury activity to advocate early controlled activity to promote healing and accelerate restoration of function. The opponents of this view argued that absolute rest allows healing to proceed at maximum pace and that early use of injured musculoskeletal structures increases inflammation and disrupts repair tissue, thus delaying or preventing healing (Buckwalter, 1995).

Although the structure and vascularity of different soft tissues (muscle, ligament, tendon) differs and the mechanism of injury may differ, the overall process which occurs at the site of injury in an attempt to heal the area, is essentially the same (Houglum, 1992). Following injury, the connective tissue is either partially or totally disrupted and a gap appears between the ruptured fibres that are retracted. The rupture gap is soon filled by a haematoma and later by proliferating granulation tissue (Jarvinen and Lehto, 1993).

Immobilisation has previously been the method of choice used in the treatment of soft tissue injuries, but due to various complaints and complications after prolonged immobilisation, early mobilisation has become more favoured as a form of treatment. Immobilisation following injury basically means complete rest until pain-free activity is feasible and can mean anything up to two to six weeks (O’Donaghue, 1976; Reider et al, 1993). Jarvinen and Lehto (1993) cited Corrigan (1967) in defining early mobilisation for muscle injuries as ‘gradually increasing physical activity involving the injured muscle within the limits of pain tolerance’. Immediate mobilisation (following muscle injury) may result in dense scar formation in the area that inhibits muscle regeneration. Immobilisation limits the area of connective tissue scarring thus allowing muscle fibre regeneration, but the orientation of the muscle fibres is haphazard and not parallel to the uninjured fibres (Jarvinen and Lehto, 1993). The same situation has been found in healing ligaments and tendons (Bair et al, 1980 and Gelberman et al, 1982, cited Buckwalter, 1995).

For those patients (particularly competitive athletes) who are reluctant to rest, some element of agreement for “relative rest” may be required to ensure compliance. Consultation will be required to ensure that healing of the injured tissues is not compromised, while “controlled activity” may be carried out.

Guideline 2 - Rest

1 - Rest should be applied to the injured part immediately following injury (II).

2 - Stress on the injured tissue should be avoided during the early (inflammatory) phase of the healing process, as the tensile strength of the injured tissue is greatly reduced at this time (II).

3 - The optimum period of rest appears to be one to five days, depending on the severity of injury. Moderate (second degree) muscle injuries require three to five days “immobilisation”. Mild (first degree) may require only 24 hours rest and severe (third degree) injuries may require at least one week’s rest (II).

4 - Early mobilisation following the period of rest should initially avoid undue stress on the healing tissue (III).

5 - Isometric work may be performed during the period of rest, within the limits of pain tolerance (C).

6 - Overall general activity should be reduced to avoid increasing metabolic rate and producing a generalised increase in blood flow (III).

Evidence

Animal studies have shown that controlled loading applied at the optimal time can promote healing of dense fibrous tissue (Gelberman et al, 1982, cited Buckwalter, 1995) and muscle tissue (Jarvinen and Lehto, 1993). Excessive early loading may increase joint instability (Burrough and Dahners, 1990) and result in excessive scarring and inhibition of muscle regeneration (Jarvinen and Lehto, 1993). A short period of immobilisation (three to five days for a second degree injury in rats) accelerates formation of a granulation tissue matrix, followed by mobilisation which promotes penetration of the connective tissue scar by muscle fibre and the orientation of regenerated muscle fibres in alignment with the uninjured fibres. Although restricted to animal studies, the results may be extrapolated to human subjects to provide a sound biological base for practice.

Clinical studies have concentrated on the mobilisation / immobilisation debate and on comparison between surgical and non-surgical methods. Few studies have given a precise indication of the duration (if any) of a period of rest following injury. Generally, clinical studies have found that immobilisation results in significant weakness and muscle atrophy, even up to five years following injury (Rutherford et al, 1990). Early mobilisation resulted in decreased need for pain medication (McCarthy et al, 1993), fewer days off work (Brooks et al, 1981), minimal treatment morbidity (Reider et al, 1993), and an earlier return to functional and sport related activity (Reider et al, 1993). Thus, clinical studies provide empirical support for early mobilisation but have yet to determine the optimal period of rest to precede mobilisation.

Recommendations for guidelines

In the absence of evidence from clinical trials regarding the precise timing of early (active) rehabilitation, evidence from animal trials, which indicates that activity during a period of up to five days following soft tissue injury may disrupt the healing process (Jarvinen and Lehto, 1993; Burrough and Dahners, 1990) forms the basis of recommendations for rest of the injured area for up to five days following injury, depending on the severity of injury.

Current practice - results of questionnaire

Of those who responded to this question (95 per cent of the total), 96 per cent stated that they prescribed rest, whilst four per cent stated that they did not. Although most of the respondents usually or sometimes prescribed complete rest of the injured part (89 per cent), 11 per cent rarely prescribe complete rest and 95 per cent prescribed activity within the limits of pain. In view of the findings by Jarvinen and Lehto (1993) that second degree injuries require three to five days’ rest, it appears that this is one area of practice that needs to be changed.

8.3 - Ice

Ice (or other applications of cryotherapy) has been recognised as one of the least expensive and widely used therapeutic modalities employed in the management of acute musculoskeletal injuries (Knight, 1989). This same author further notes that the confusion surrounding cryotherapy is almost as extensive as its popularity, with different opinions existing concerning its theoretical base, techniques of application and the physiological responses of the body to its application.

Claims for the effects of cryotherapy include decrease in pain, decrease in metabolism, decrease in swelling, decrease in muscle spasm, decrease in circulation (but also cold-induced vasodilation) and effects on the inflammatory process. In spite of the fact that some of these claims are unsubstantiated, it appears that they still form the basis for the application of ice during the rehabilitation process.

Some studies demonstrate an initial increase in pain upon application of ice, followed by decrease in pain and numbness (Wolf and Hardy, 1941, cited Knight, 1989), and a cyclical course of pain and numbness (Kunkle, 1949, cited Knight, 1989). Clinical as well as experimental research on pain and the pain threshold indicates that pain reduction occurs after cooling to temperatures around 10- 15 0 C, although uncertainty exists as to the duration of the analgesic effect (Meeusen and Lievens, 1986). Lee et al (1978) found reduction in both motor and sensory nerve conduction velocity in the ulnar nerve during application of ice. Knight and Londeree (1980) showed decreased blood flow during cold applications. However, they also found that exercise appeared to supersede the effects of cold on blood flow, indicating that even in the presence of cold application, exercise will increase blood flow. Taber et al (1992) demonstrated a reduction in the normal increase in local blood volume during venous occlusion, when a cold gel pack was applied to the ankle, with no apparent reactive cold-induced vasodilation.

Decreased metabolism is the primary reason for using cryotherapy during the immediate care of acute musculoskeletal injury (Knight, 1989; McLean, 1989; Rivenburgh, 1992). Hypothermia reduces cellular energy needs, thereby reducing the need for and conserving limited supplies of oxygen (Knight, 1989).

Research into the effect of application of cold on oedema has produced conflicting reports. This was summarised by Kowal (1983), who stated that ‘the right kind of cold can be effective in decreasing swelling…the wrong kind of cold can exacerbate swelling.’ Results from clinical studies are complicated by the fact that in first aid, cold is usually combined with compression (Meeusen and Lievens, 1986).

Clinically it is suggested that appropriate application of therapeutic heat and cold encourages rehabilitation (Kaul and Herring, 1994), minimises the acute response to soft tissue injury and maximises recovery from chronic sequelae (Halvorson, 1990).

Perhaps because of its easy accessibility and general acceptance as a treatment modality, it is easy to assume that the application of therapeutic cold has few dangers. Clinical observations by some authors suggest that care must be taken when applying cold over superficial nerves (Covington and Basset, 1993; Basset et al, 1992; Green et al, 1989). Although not supported in the literature, clinical observations suggest that the static application of cold (in the form of ice or commercial cold packs) may result in what is commonly termed an ‘ice burn’. This may be regarded as a form of superficial frostbite, in which tissue destruction results from exposure to low environmental temperature. The signs and symptoms are similar to a thermal burn, with pain, swelling, redness and blistering.

Guideline 3 - Ice

1 - Ice should be applied immediately following acute musculoskeletal injury (C).

2 - Chipped / crushed ice in a damp towel appears to be the most effective application of cold (II) followed by ice in a plastic bag, and then frozen gel packs (II).

3 - A damp towel should be placed between the cooling agent and the skin to avoid “ice burn”. A maximum safe period for icing of 30 minutes is recommended, to avoid skin and tissue damage and nerve palsy (C).

4 - The most effective duration of application of ice appears to be 20-30 minutes, applied every two hours, although there are no specific recommendations (II).

5 - Areas with >2cm subcutaneous fat may require longer applications (30 mins) since it has been found that ten minutes of application produces no muscular cooling effect in these circumstances (III).

6 - Cold application should cover the entire area affected by the injury (III).

7 - The athlete should not return to participation immediately following application of ice (or other types of cold application), as nerve conduction velocity, sensation and connective tissue flexibility are likely to be reduced by cold application (C).

8 - Care should be taken in the application of ice to patients with little subcutaneous fat and in the region of superficial nerves, e.g. common peroneal nerve at the knee, ulnar nerve at the elbow, since cold-induced nerve injury may result. To compensate for this, the duration of application should be reduced (no more than ten minutes) or an insulating material should be applied between the source of cooling and the patient’s skin (III).

9 - Application of cold is contraindicated for patients who have previously developed cold-induced hypertension during cold treatment, who have allergy to cold (urticaria, joint pain) or who have Raynaud’s syndrome, peripheral vascular disease or sickle cell anaemia (C).

10 - If nerve damage as a result of the injury is suspected or if there is a history of reduced skin sensation, application of cold should not exceed 20 minutes and the skin condition should be checked every five minutes (C).

Evidence

The local application of cold (cryotherapy) was mentioned by Hippocrates (460-370 BC), who advised the use of ice and snow as a therapeutic modality. Meeusen and Lievens (1986) and Ogilvie-Harris and Gilbert (1995) list some of the effects attributed to cryotherapy, i.e. limiting the inflammatory response, reduction of oedema, reduced haematoma formation and pain relief. Physiological responses such as vasoconstriction, decrease of blood flow, reduction of muscle spasm and decrease in nerve conduction velocity have also been attributed to cryotherapy.

The use of cold modalities to decrease the effects of a local inflammatory response is widely accepted. Tissue cooling produces an initial vasoconstriction in the cutaneous blood vessels, resulting in a decrease in local blood volume. Controversy has surrounded the subsequent responses, since Lewis (1930, cited Kowal, 1983) first described cyclic phases of skin cooling and rewarming (the ‘hunting response’) when tissue temperature reached 15 0 C when a subject’s finger was immersed in ice water. Knight and Londeree (1980) measured blood flow using strain gauge plethysmography in 12 normal subjects during therapeutic applications of heat, cold and exercise. They found that total blood flow was greater (p<0.002) during a combination of cold and exercise than during application of heat. However, there was neither cold-induced vasodilation nor a reflex vasodilation following 25 minutes of cold application, suggesting that during cryotherapy it is exercise that causes the increased blood flow rather than the application of cold.

Taber et al (1992) measured blood flow in 13 normal subjects, under three experimental conditions of rest, application of a room temperature gel pack and application of a cold gel pack. A significant reduction in local blood volume was found for the cold gel pack, with the maximum decrease occurring 13.5 minutes after application. No reactive vasodilation was observed. Kowal (1983) and Meeusen and Lievens (1986) report inconsistent findings from a number of studies on the effects of cold application on blood flow, suggesting that the differences in outcome are a result of the variation in methodologies used. However, it appears that reactive vasodilation with resultant increase in blood flow occurs at skin temperatures below 14 0 C, whereas at skin temperatures between 14 0 C and 42 0 C, vasoconstriction and decreased blood flow occurred. Thus empirical evidence seems to support vasoconstriction and decrease in blood flow at temperatures between 14 0 C and 42 0 C.

It is difficult to find definitive recommendations on the frequency and duration of cold application in the literature. Ho et al (1994) noted that few studies had examined the effects of cold on blood flow over time, and that those which had were performed on arms and fingers, and had produced no consensus. These researchers found decreases in soft tissue blood flow and skeletal blood flow with as little as five minutes ice application, which effects were enhanced three- to four-fold by increasing the duration of application to 25 minutes. This study supported earlier work by Knight and Londeree (1980), who also applied ice for 25 minutes. Belitsky et al (1987) found skin cooling of up to 12 0 C after 15 minutes application, with skin temperatures rising towards their pre-ice levels 15 minutes following removal of the ice. The effect of longer durations of application has not been established. Recommended frequency of application varies between 20 minutes in every hour (McMaster et al, 1978) to 30 minutes every two hours (Knight, 1989).

Contrary to commonly held beliefs, it appears that cold application does not reduce oedema but may indeed result in increased swelling. Farry and Prentice (1980) studied the effects of ice treatment on experimentally induced ligament injury in the domestic pig. They found that there was greater swelling in the ice-treated injured limbs, but even in uninjured animals, there was greater swelling in those having ice application, than in the controls (no injury, no ice). The temperature of the ice was not identified in this study. Other studies cited by Meeusen and Lievens (1986) report inconsistent findings. However, those studies in which temperature was quoted suggest that lower temperatures (<15 0 C) seem to result in increased oedema when compared with limbs exposed to temperatures of >30 0 C. Prolonged exposure to less intense cold may also be implicated (McMaster et al, 1978). Meeusen and Lievens (1986) suggest that the reasons for the increased oedema is increased permeability of the lymph vessels, with resulting extravascular protein accumulation and increased extravasated fluid. Laba (1989) found no difference in swelling between an experimental group with ankle injuries who received ice pack treatment, and a control group which did not. Thus empirical and biological evidence from animal and human studies seem to refute the notion that cold application reduces oedema. The same evidence is not widely apparent from clinical studies, but this may be because cold application is usually combined with compression and elevation (Farry and Prentice, 1980).

Kowal (1983) outlined several reports that have demonstrated the pain-relieving capacity of treatments involving cold application. He cited a study by Chambers (1969), who applied cold in the form of towels soaked in ice water to 23 patients with a variety of musculoskeletal and neurological conditions. He found that pain was sufficiently decreased or eliminated in 74 per cent of those patients with pain. Meeusen and Lievens (1986) also stated that clinical, as well as experimental, research on pain and the pain threshold indicates that pain reduction occurs after cooling to temperatures around 10-15 0 C, although uncertainty exists as to the duration of the analgesic effect. Nevertheless, Lee et al (1978) found that application of ice for between 16 and 24 minutes resulted in a significant reduction in both sensory (p<0.005) and motor (p<0.005) nerve conduction velocity which may explain the reduction in pain. Most of the findings on the pain-relieving ability of cold application are based on empirical, clinical evidence but there is a lack of evidence from controlled trials comparing the mode or duration of application, the ideal temperature, and the duration of the analgesic effects. This lack of evidence regarding the mode of application of ice (or cryotherapy) has been noted in the literature (Ho et al, 1994; McDowell et al, 1994).

Although the application of cold is regarded as a readily accessible, inexpensive and safe modality, case reports indicate that there is a risk of nerve damage if the application is prolonged, or over superficial nerve trunks, or in athletes who have little subcutaneous fat. Covington and Basset (1993) and Basset et al (1992) reported six case studies in which peripheral nerve injury followed application of ice, and Green et al (1989) also reported a case study involving peripheral nerve injury following ice application. The nerve most commonly affected was the common peroneal nerve following ice application over the lateral aspect of the knee (four cases), although the lateral femoral cutaneous nerve was implicated in two cases when ice had been applied over the anterior superior iliac spine. The supraclavicular nerve was affected in one case following ice application over the shoulder. Durations of application varied between 20 minutes and one hour. The precise mode of application of ice was noted in only two cases (plastic ice bag and ice bag over a towel). This evidence, obtained from clinical observations, suggest that caution is required when applying ice in close proximity to superficial nerves.

In subjects with large subcutaneous deposits of fat, cooling may be impaired (Wolf and Basmajian, 1973). In a study in which deep muscle temperatures were recorded in the gastrocnemius muscle in ten subjects over a period of five minutes, the subject who was described as being 'of a muscular body type' demonstrated the greatest temperature change, whilst the subject who was described as 'obese' demonstrated the smallest temperature change. Although the sample size in this study was not large enough to draw definitive conclusions regarding the influence of obesity on cooling, it appears that longer periods of cooling may be required in obese individuals to achieve a therapeutic effect.

Meeusen and Lievens (1986) have identified several modes of application of cold. The most common therapeutic application is usually with ice packs, ice towels or ice massage to the injured part (Kalenak et al, 1975 and Kern, 1980, cited Meeusen and Lievens, 1986). Another application technique, mostly used in research because of the ability to control temperature, is immersion in cold water. Local cryotherapy can also be applied by frozen gel packs, vapocoolant sprays, chemical ice and refrigerant inflated bladders / splints. Skin temperature has been reduced by between 6 0 C (ice pack) and 29.5 0 C (water immersion at 4 0 C) using different applications, applied for between ten seconds (vapocoolant spray) and 193 minutes (water immersion at 4 0 C). Intramuscular temperatures show a delayed fall in temperature that continues after the cold application has been removed (Meeusen and Lievens, 1986).

Few comparative studies have been carried out to evaluate the efficacy of different cold applications. However, McMaster et al (1978) compared chipped ice, frozen gel packs, chemical ice pack and an inflatable envelope containing a gaseous refrigerant, measuring deep intramuscular temperatures over a period of one hour, in canine thighs. The iced gel packs performed best of the artificial ice techniques but the chipped ice performed best overall, producing a temperature drop of 11.3 0 C. Belitsky et al (1987) also found crushed ice in a damp towel (wet ice) to be the most effective application, performing better than crushed ice in a plastic bag (dry ice) or cryogen packs, both in the actual temperature drop, and the duration of effect. These authors also found that the direct effects of the cold application were restricted to the area of application. Botte (1982, cited Meeusen and Lievens, 1986) found no depth effect with ethyl chloride sprays. Thus, although limited, empirical evidence tends to support chipped ice as the most effective form of cold application.

Scheffler et al (1992) support our findings that no specific conclusions have been reached regarding the optimum duration of cold applications, possibly due to different outcome measures being used.

Recommendations for guidelines

However, although the evidence is not conclusive, there is general agreement that the application of ice (or cold) for 20-30 minutes results in decreased pain, blood flow and metabolism (Taber et al, 1992, Ho et al, 1994, McMaster et al, 1978; Knight, 1989). Duration of application should be reduced over superficial nerve trunks, or in patients with little subcutaneous fat (Covington and Basset, 1993; Basset et al, 1992), but increased in individuals with large subcutaneous deposits of fat (Wolf and Basmajian, 1973). The most effective mode of application of cold appears to be chipped ice in a damp towel (McMaster et al, 1978; Belitsky et al, 1987).

Current practice - results of questionnaire

Crushed ice was the most common mode of application of cold in the hospital/ clinic setting (92 per cent of respondents usually or sometimes using this application), with gel packs and cryocuff usually or sometimes being used by 76 per cent and 65 per cent of respondents respectively. Only 21 per cent usually or sometimes used cold sprays in this setting. Crushed ice, gel packs and ice cubes were the most common modes of application in the sporting environment, with 86 per cent of respondents usually or sometimes using these applications. Cold sprays were more popular in this environment with 75 per cent usually or sometimes using these. Damp towels were the most popular mode of application of ice in the hospital/ clinic setting, whereas in the sporting environment, plastic bags, damp towels and a combination of plastic bags + damp towels were used with equal frequency. Most respondents applied ice/ cold for between 11 and 20 minutes (47 per cent), with a frequency of three to five times per day. However, 41 per cent applied ice / cold for more than 20 minutes and 28 per cent suggested its application hourly when awake.

8.4 - Compression

Poor management of acute musculoskeletal injuries can lead to excessive swelling, which in turn may develop into chronic oedema. Following knee arthroplasty, it has been found that discomfort and oedema can lead to inhibition of the extensor mechanism, with delayed resumption of motion, strength and leg control (Whitelaw et al, 1995). Although the application of compression has long been accepted as a means of limiting and reducing oedema, most of the support for the use of compression dressings has arisen from observation and convention. Ice, compression and elevation are widely recognised as the standard intervention in acute soft tissue injuries, and it is difficult to find studies in which the effects of compression alone have been investigated.

External compression through the application of an elastic wrap can stop bleeding, inhibit seepage into underlying tissue spaces and help disperse excess fluid (Thorsson et al, 1997). As the compression increases the hydrostatic pressure of the interstitial fluid, this fluid is pushed back into the capillaries and lymph vessels in the region of the trauma or into the tissue spaces away from the traumatised area (Rucinski et al, 1991). The presence of external compression increases the effectiveness of the muscle pump in influencing venous return.

Several modes of application of compression are available, ranging from non-adhesive and adhesive elastic bandages and tubigrip to adjustable neoprene supports and inflatable splints.

Guideline 4 - Compression

1 - Always apply compression in a direction from distal to proximal, irrespective of the type of bandage/ compression agent (C).

2 - Pressure must not be greater proximally than distally. Application of pressure should be uniform throughout the compression (II).

3 - Apply compression a minimum of six inches below and above the site of injury. At distal sites (e.g. ankle, wrist) apply from heads of metatarsals/ metacarpals to the joint proximal to the site of the injury (C).

4 - Apply as per manufacturer’s instructions when these are available (C).

5 - Compression must be capable of accommodating oedema immediately following injury, and of continuing to apply pressure with diminishing effusion (II). Therefore:

• Do not apply compression with the material at full stretch
• Ensure overlap of half to two-thirds of previous turn of compressive material
• Apply turns in a spiral fashion - never apply circumferential turns
• Protective padding (gauze, underwrap, foam, cotton wool, gamgee) or gapping of the compressive material may be necessary over vulnerable areas such as the popliteal fossa, superficial tendons and bony prominences

6 - Compression using an elastic legging (or equivalent) should not be applied in the recumbent (lying) position, or in association with elevation (I).

7 - Remove and reapply if uniform and constant pressure is not maintained, or to administer other treatment modalities. Otherwise, reapplication is recommended within 24 hours (C).

8 - Compression should be applied as soon as possible following injury (C).

9 - Intermittent compression may be applied (30 mins daily at a compression of 60mmHg, 30 seconds on, 30 seconds off) in addition to compression during the first five days (II).

10 - Continue compression for first 72 hours following injury, when not lying down (II).

11 - Distal areas should be checked immediately following application of compression for signs of diminished circulation (cold, pallor) and then regularly checked throughout the continued application of compression (C).

12 - The following materials may be considered for purposes of application of compression:

• Cohesive / tensor bandages
• Tubigrip
• Elastic adhesive bandage
• Adjustable neoprene supports
• Inflatable pressure devices. Elastic bandages and tubigrip appear to be most effective (II)

Evidence

Few studies exist which examine the effects of compression alone on oedema or in the management of acute soft tissue injuries, as most seem to investigate the influence of a combination of cold application with compression (Schroder and Passler, 1994).

A study by Rucinski et al, (1991) compared three treatment protocols in the management of post-traumatic oedema in 30 subjects following lateral ligament ankle sprains. Subjects were divided into three groups. One group had the limb elevated at 45 0 C for 30 minutes, the second had the limb elevated at 45 0 C and an elastic wrap applied from the heads of the metatarsals to 12.7 centimetres above the malleoli for 30 minutes, and the third group had the limb elevated to 45 0 C and an intermittent pressure device applied at 40-50mmHg for 30 minutes. Both of the compression protocols resulted in slightly increased limb volume when measured by water displacement and the elevation alone resulted in a significantly decreased limb volume, suggesting that elevation alone is the treatment of choice. However, although elevation alone produced the most favourable results, in reality this is difficult to maintain over prolonged periods of time and it may be that compression has its most beneficial effect when the limb is not elevated. Furthermore, the degree of compression (40-50mmHg in the intermittant compression device) may have caused arteriolar dilatation under the area of compression (Neilsen, 1983b), which on removal of the compressive device caused increased blood flow and increase in limb volume.

Airaksinen et al (1990) compared the effectiveness of elastic bandages and intermittent compression for the treatment of acute ankle sprains. Subjects who had sustained an inversion injury to the ankle were randomised into two groups, the control group wore elastic bandages and the experimental group hade intermittant compression therapy for once per day for five days in addition to wearing elastic bandages. At one week and four week follow-up measurements, both groups demonstrated improvements in all measures, but the compression group had significantly less oedema, less pain, greater range of movement and improved function than the control group. This suggests that a combination of elastic bandage and intermittent compression is better than an elastic bandage alone. Although the study was in the form of a randomised controlled trial, no information was given about advice on elevation, weight-bearing or the position in which the compression treatment was given (recumbent or elevated), potentially reducing the scientific rigour of the report.

Murthy et al (1994) provided both support and a possible explanation for Rucinski et al’s (1991) findings; they found that in the recumbent position, elastic leggings created intramuscular pressures of a magnitude which had the potential to compromise microcirculation, and suggested that patients should be discouraged from lying down with elastic leggings for long periods of time. The authors did not identify the period of time, but as the reported pressures in the study were recorded at 30 seconds, it might be assumed that lying down with elastic leggings should be avoided completely. Lower “safe” intramuscular pressures were recorded during sitting, standing, walking and running.

Thorsson et al (1997) in a prospective non-randomised trial of 40 athletes, found no difference in range of motion, serum creatine kinase and ultrasonic scan following muscle injuries between two groups, one of which had immediate (within five minutes) compression and the other which had rest and elevation only. The lack of difference between the two groups may be explained by the fact that the compression group appeared to have continuous compression, which other studies have suggested may be undesirable, until symptom-free.

Neilsen (1983b) applied external compression at levels of 10mmHg up to the level of diastolic pressure of 66-74mmHg. At external pressures of greater than 10mmHg, blood flow in both subcutaneous and muscle tissue was significantly reduced (p<0.05). However, Neilsen (1983b) suggested that increased tissue pressure provides a stimulus for arteriolar dilatation as a means of autoregulation of blood flow, but that this compensatory dilatation is masked by a general increase in local vascular resistance caused by the external pressure. This study provided empirical evidence for reduction of blood flow during compression and the biological basis for the increase in limb volume when compression is removed.

Furthermore, Neilsen (1983b) showed that when proximal pressure was applied at a level of 66- 70mmHg, distal external pressures of 10-40mmHg resulted in significant (p<0.005) increases in subcutaneous blood flow and small but insignificant (p>0.01) increases in muscular blood flow. This provides empirical evidence to support avoidance of high proximal-distal pressure gradients.

Matsen and Krugmire (1974, cited Brodell et al, 1986) studied tibial fractures in the rabbit and concluded that although external pressure decreased post-fracture swelling, the pressure must be uniform and controlled so that it does not result in adverse haemodynamic effects. They observed that most pressure dressings currently in use tend to apply more pressure as the volume of the limb increases, with undesirable consequences. This provides empirical support for careful grading of applied compression which must accommodate increases in limb volume brought about by oedema. Precise pressure levels were not quoted.

Brodell et al, (1986) showed that the application of compression in the form of a Robert Jones bandage, with a distal to proximal pressure gradient resulted in reduced antero-lateral compartment pressure which was reversed upon removal of the bandage. Tufft and Leaman (1994) found that immediately following application, a wool and crepe bandage produced more pressure than an elastic tubular bandage at the ankle of both normal and ankle-injured subjects, but suggested that over time the tubular bandage had greater effect.

Numerous studies have evaluated the effectiveness of a combination of ice/cold and compression on pain and oedema in a variety of clinical situations (Starkey, 1976; Quillen and Rouillier, 1982; Scheffler et al, 1992; Levy and Marmar, 1993; Healy et al, 1994; Whitelaw et al, 1995; Barlas et al, 1996). Due to considerable variation in study design and outcome measures, direct comparison across studies is not possible. However, general conclusions are that the combination of ice and compression reduces pain and swelling. Thus these studies provide empirical evidence for the efficacy of a combination of ice and compression but do not distinguish between the two.

Recommendations for guidelines

The findings from these studies indicate that the application of compression can effectively reduce oedema and pain, and improve range of motion and function following soft tissue injury (Rucinski et al, 1991; Airaksinen et al, 1990). Evidence also suggests that compression should be removed when the limb is elevated, but should be applied at all times when the limb is in the dependent position (Rucinski et al, 1991; Murthy et al, 1994). Additionally, intermittant compression may be of value in the early stages (Airaksinen et al, 1990), but not in the elevated position. Applied pressure should be controlled with a distal to proximal pressure gradient (Brodell et al, 1986).

Current practice - results of questionnaire

Ninety-eight percent of the total sample stated that they prescribed compression , which was 100 per cent of those who responded to this question. Tubigrip and taping / strapping were the most popular modes of application in both the clinical and sporting environments, with stretch bandages being the third most popular in both locations. Crepe bandages were used (usually or sometimes) by 36 per cent (clinic) and 33 per cent (sporting environment) of the total sample. This perhaps gives some cause for concern due to the material’s limited ability to accommodate to increasing limb volume. Similarly, one might question the small number (11 per cent in the clinic and 12 per cent in the sporting environment) who stated they used non-stretch bandages.

The majority of respondents always or sometimes advocated removal of compression for application of ice / cold (94 per cent), other treatment (96 per cent) and at night, whilst 66 per cent advocated the permanent application of compression. Apparent discrepancies in numbers are because the total sample did not respond to all questions.

8.5 - Elevation

Kellet (1986) noted that it is common practice to combine cryotherapy with both compression and elevation of the injured limb to reduce the effects of the local inflammatory response. He further suggested that, although specific data are not available to support these elements of treatment, logic suggested that elevation assists in overcoming the gravitational influence on the accumulation of oedema.

Guideline 5 - Elevation

1 - Elevate the injured part above the level of the heart as much as possible during the first 72 hours following injury (III).

2 - Ensure that the elevated part is adequately supported (e.g. by pillows, slings) (C).

3 - Elevate the injured part as soon as possible following injury (C).

4 - Avoid placing the limb in the dependent position immediately following elevation as the ‘rebound phenomenon’ will tend to increase oedema (II).

5 - If the limb can be maintained in elevation, do not apply compression simultaneously (I).

Evidence

Neilsen (1983a) found that intra-arterial pressure reduction was achieved by limb elevation above heart level in subjects placed in a horizontal position. This pressure reduction has the further effect of providing a stimulus for arteriolar dilatation, similar to that resulting from increased tissue pressure produced by external compression. Baumert (1995) noted that elevation of the injured part above the level of the heart enhances draining of extravascular fluid away from the injured area, but does not support this observation with evidence. From the Neilsen (1983a) study, reduction in intra-arterial pressure resulting from elevation of the limb might cause a reduction in extravasation of fluid into the interstitial tissues, suggesting but not confirming a biological base to support elevation of the injured part.

Empirical evidence to support the use of elevation was produced by Rucinski et al (1991), who found that elevation alone was more effective in reducing oedema following ankle injury than elevation plus intermittant compression or elevation plus elastic wrap. As the measurement of oedema was by volumetric displacement, the superior effects of elevation alone may be explained by greater arteriolar dilatation when the tissues were also compressed which resulted in an increase in limb volume once the compression had been removed (Neilson, 1983b). It may also be explained by the ‘rebound phenomenon’, which is caused by a sudden shift of vascular or lymphatic fluid when the limb is moved from an elevated position to a dependent position. This does not occur with elevation alone, but only when elevation is combined with compression. Furthermore, it might be suggested that if compression is applied to an elevated limb, arterial supply and venous drainage may be compromised.

Recommendations for guidelines

This evidence tends to suggest that when an injured limb / part can be maintained in elevation above the level of the heart, compression should not be applied simultaneously (Rucinski et al, 1991). However, when elevation cannot be maintained, evidence suggests that compression reduces intra-arterial pressure and increases tissue pressure, thus reducing extravasation of fluid into the tissues (Neilsen, 1983b).

Current practice - results of questionnaire

Ninety-eight per cent of the total sample stated that they prescribed elevation, and of these, 91 per cent advocated elevation ‘as much as possible’ and 59 per cent suggested continual elevation. Fifty-six per cent of the total sample prescribed elevation of the injured part above the proximal joint, 53 per cent prescribed elevation above the heart and 48 per cent prescribed supine lying with elevation above the level of the heart.