5.2. CLINICAL ASSESSMENT IN THE MANAGEMENT OF DETOXIFICATION
5.2.1. Clinical assessment of dependence
Most service users presenting for detoxification will show a clear history of opioid dependence, whether by being on prescribed methadone or buprenorphine, or by the clinical presentation of signs of illicit heroin use (for example, abundance of needle marks). Some may have been misusing other opioids additional to any prescribed medication. Often service users may also misuse and be dependent on benzodiazepines and/or alcohol, or stimulants such as cocaine or amphetamines.
It is important that any opioid detoxification regimen should be appropriate to the service user’s degree of dependence and the extent of the withdrawal symptoms he or she experiences. Errors have occurred where service users have persuaded the health-care professional conducting a clinical assessment that their degree of opioid use and/or dependence is significantly greater than it is in reality; in some such cases they have had no dependence on or even use of opioid drugs at all. This can lead to the prescription of dangerously high doses of opioids. Adequate assessment of a service user’s opioid dependence status is therefore crucial prior to undertaking opioid detoxification.
Opioid dependence is normally diagnosed primarily through a clinical assessment but can be assisted by testing for drugs in biological fluids and by the use of psychometric measures. The clinical assessment of opioid dependence involves asking the service user about the pattern and nature of his or her drug use, the extent of use and treatment episodes in the past, to ascertain the degree of dependence (DH, 1999). A formal psychometric measure may sometimes be employed as an aid to the assessment of dependence. For example, dependence is diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) when three or more of the following criteria are present in a 12-month period: tolerance; withdrawal; increasing use over time; persistent or unsuccessful attempts to reduce use; preoccupation or excessive time spent on use or recovery from use; negative impact on social, occupational or recreational activity; and continued use despite evidence of its causing psychological or physical problems (APA, 1994).
The use of biological testing is important to confirm the reported use of specific drugs, including prescribed and illicit opioids and other non-opioid drugs. In addition, an examination of physical and psychiatric health is important to assist diagnosis of dependence and to assess any further complication to the process, such as comorbid physical or mental health problems or pregnancy (DH, 1999).
The clinical assessment of opioid dependence aids the clinician in determining the level of caution required during detoxification. In particular, if the service user has a low level of dependence or uncertain tolerance, it is vital that detoxification is conducted in a setting that allows the clinician to observe withdrawal symptoms and titrate medication accordingly. In general, detoxification is not required for people who misuse drugs but are not dependent. In addition, caution is also required where polysubstance use or possible polysubstance dependence (commonly alcohol and benzodiazepines) is detected. Polysubstance dependence can complicate the detoxification process and settings for titration therefore need to be appropriate for the level of observation required.
Where a clinical assessment determines that the service user is misusing alcohol, in addition to being opioid dependent, attempts should be made to address this. The possibility should also be noted that a service user may substitute alcohol for his or her previous opioid misuse during or after the detoxification process. Where alcohol dependence is present, detoxification of alcohol should also be considered either before (in community-based settings) or, if there is adequate medical supervision (for example, inpatient settings), concurrently with opioid detoxification.
If a service user is dependent on benzodiazepines, the severity of dependence and the preference of the service user should be taken into account when deciding whether to detoxify from benzodiazepines concurrently or separately from opioids.
5.2.2. Clinical assessment and monitoring of withdrawal
It is important to assess both objective and subjective withdrawal symptoms, at the start of treatment and during the induction and withdrawal stages. This is necessary in order to titrate the medication to alleviate withdrawal symptoms (DH, 1999). The objective signs of withdrawal can be assessed through careful monitoring of the service user’s pulse, blood pressure, agitation and sedation. In addition, asking the service user about the subjective signs of distress should also form part of the assessment. Formal psychometric tools may be useful in that they aid standardisation, but they are not a substitute for appropriate clinical assessment. Regular review is crucial because an overdose of methadone during detoxification may initially present as sedation and/or sleepiness, with under dosing presenting as agitation and anxiety.
5.2.3. Clinical practice recommendations
Clinical assessment of dependence
- 5.2.3.1.
People presenting for opioid detoxification should be assessed to establish the presence and severity of opioid dependence, as well as misuse of and/or dependence on other substances, including alcohol, benzodiazepines and stimulants. As part of the assessment, healthcare professionals should:
use urinalysis to aid identification of the use of opioids and other substances; consideration may also be given to other near-patient testing methods such as oral fluid and/or breath testing
clinically assess signs of opioid withdrawal where present (the use of formal rating scales may be considered as an adjunct to, but not a substitute for, clinical assessment)
take a history of drug and alcohol misuse and any treatment, including previous attempts at detoxification, for these problems
review current and previous physical and mental health problems, and any treatment for these
consider the risks of self-harm, loss of opioid tolerance and the misuse of drugs or alcohol as a response to opioid withdrawal symptoms
consider the person’s current social and personal circumstances, including employment and financial status, living arrangements, social support and criminal activity
consider the impact of drug misuse on family members and any dependants
develop strategies to reduce the risk of relapse, taking into account the person’s support network.
- 5.2.3.2.
For women who are opioid dependent during pregnancy, detoxification should only be undertaken with caution.
- 5.2.3.3.
For people who are opioid dependent and have comorbid physical or mental health problems, these problems should be treated alongside the opioid dependence, in line with relevant NICE guidance where available.
Care for people who misuse other medicines and/or substances in addition to opioids
- 5.2.3.4.
If a person presenting for opioid detoxification also misuses alcohol, healthcare professionals should consider the following.
If the person is not alcohol dependent, attempts should be made to address their alcohol misuse, because they may increase this as a response to opioid withdrawal symptoms, or substitute alcohol for their previous opioid misuse.
If the person is alcohol dependent, alcohol detoxification should be offered. This should be carried out before starting opioid detoxification in a community or prison setting, but may be carried out concurrently with opioid detoxification in an inpatient setting or with stabilisation in a community setting.
- 5.2.3.5.
If a person presenting for opioid detoxification is also benzodiazepine dependent, healthcare professionals should consider benzodiazepine detoxification. When deciding whether this should be carried out concurrently with, or separately from, opioid detoxification, healthcare professionals should take into account the person’s preference and the severity of dependence for both substances.
5.3. DRUG TESTING
5.3.1. Introduction
The analysis of human body fluids can yield important information in support of healthcare professionals’ caring for service users who are about to undertake, or who are undertaking, opioid detoxification. Such analyses are only an adjunct to an appropriate clinical investigation of the service user. Currently, no single test is available that is able to establish or confirm a diagnosis of drug dependence.
In drug misuse services, oral fluid or urine testing are commonly employed, while hair and blood testing are utilised to a lesser extent (NACB, 2006). The numerous testing procedures available can provide evidence of drug consumption, trend of use over time when repeated, and compliance with prescribed drugs.
Moreover, testing may also be useful during a longer-term detoxification, to assess compliance with prescribed medication and to ascertain possible use of illicit drugs. Random intermittent interval testing is probably the most clinically and cost-effective regime. It will help the clinician in confirming the clinical picture and aid assessment of the success of detoxification and possible need to review dosage.
Testing occurs in a variety of settings, including specialist drug services, primary care, residential units, prisons and some hospital settings. The rationale for testing is to help confirm opioid use and to assess other complicating factors, as well as monitoring of care. Testing can be conducted at point of care (that is, near-patient testing) or can be confirmed in a laboratory. Both forms of testing are important tools in clinical practice and will be considered in the sections below.
5.3.2. Near-patient testing
Near-patient testing refers to the process of obtaining a biological sample from a service user and using a drug-testing kit to detect immediately the presence of any of a variety of substances (for example, opioids, amphetamines, cocaine metabolite, benzodiazepines, methadone and cannabis) on site. This process eliminates the need for external laboratory support and provides rapid results.
In current practice, oral swabs or urine screening kits are most commonly used for near-patient testing. These forms of testing are used for a variety of reasons, including monitoring within a criminal justice order, arrest referral schemes, prison systems and medicolegal situations.
Current rapid screening of biological samples for misused drugs depends on immunochemical techniques. Essentially, antibodies with a specific and high affinity for a particular drug, and/or its metabolites, react with the drug present in the sample. The extent to which the antibodies have become bound to drugs present in the sample is then detected by one of several different techniques. All immunochemical methods have problems in relation to specificity, whereby the antibody employed may react with compounds in the sample other than those that the test is intended to measure (DH, in press). There are also potential issues with matrix effects, whereby problems with the sample may destroy the drug/metabolite or the antibody, or interfere with the reaction between the two.
While new technologies based on techniques such as Fourier transform infrared spectroscopy and nanotechnology are under active development and techniques using liquid chromatography in combination with tandem mass spectroscopy are starting to come into use in the laboratory, for the next 2–3 years immunochemical techniques are likely to be the basis of most rapid screening inside the laboratory or at the point of service-user contact.
The analytical, quality and safety issues involved with near-patient testing are well known to clinical laboratories (George & Braithwaite, 2002). For example, false positives may result where the identification of a specific substance may be due to the presence of artefacts or compounds in the biological matrix that are similar to the drug of interest (NACB, 2006). False positive results may also occur due to misinterpretation of a test result. The presence of morphine in urine is often assumed to be indicative of heroin use but may also reflect the consumption of analgesic preparations or poppy seeds (Mule & Casella, 1988).
The problems involved with ensuring results obtained with tests undertaken outside of the laboratory, such as pregnancy or blood glucose testing, are fit for purpose have been well described (George & Braithwaite, 2002). For example, when urine dipsticks are used, colour change must be detected to indicate the presence of an illicit substance; however, this can be difficult for the inexperienced eye (George & Braithwaite, 2002) and such processes are highly subjective. Samples must also be kept in adequate conditions, as they are susceptible to contamination. Some testing kits are only able to determine whether a drug is present but not the type or quantity.
Training and meticulous attention to the manufacturer’s instructions are essential for test results to match the levels of performance (for example, sensitivity and specificity) found in validation studies. Further, experience with other analytes measured outside the laboratory suggests the necessity for continued training of staff and the need for the use of quality assurance techniques. Where service users are being assessed in a clinic within a district general hospital, it is arguable that there is no need for near-patient testing of urine samples.
Urinalysis
Urinalysis remains the most reliable tool for identifying drug use in a drug using population (DH, in press). A further advantage of this testing method is that it can detect drug use during the previous few days. Most opioids can be detected between 2 and 3 days after use, methadone up to 9 days and cannabis up to 27 days after use (DH, 1999). However, caution must be exercised when interpreting results of urinalysis as there are a number of products commercially available specifically designed to produce false negative urinalysis results by seeking to remove illicit drugs from the body (NACB, 2006). These substances have the ability to either dilute urine samples or partially eliminate drugs, thereby making detection of illicit drugs difficult.
A recent targeted screening study by Tomaszewski and colleagues (2005) in a US emergency department found promising sensitivity and specificity for near-patient urine testing for opioids (sensitivity = 100%, specificity = 98.7%) and cocaine use (sensitivity = 96.8%, specificity = 100%), but lower sensitivity for cannabis use (sensitivity = 87.5%, specificity = 99.3%) when a comparison was made with confirmatory laboratory tests.
However, lower levels of sensitivity and specificity have been reported elsewhere. This is illustrated by the experience of the prison service, where urine samples for mandatory drug testing are collected under a high degree of supervision. On average, of all samples submitted where a screening test had produced a positive result, the confirmation test, using definitive analytical procedures such as gas chromatography/mass spectroscopy, or liquid chromatography/mass spectroscopy, did not confirm the positive screening test on 11% of occasions (HM Prison Service, 2005). In the case of opioids, only 90% of positive tests on screening were confirmed to be positive by definitive testing; for benzodiazepines this was 70%, for methadone 80% and for amphetamines 50% (HM Prison Service, 2005). It should be noted that screening tests on samples submitted for mandatory testing in prison are carried out in the laboratory using sophisticated analytical equipment rather than with kits at the point of contact.
Oral fluid testing
The major advantages of oral fluid drug testing are that it can potentially be relatively easily obtained and is less intrusive than urinalysis. It is also less open to adulteration. These properties enable oral fluid testing to be conducted by personnel with relatively little training, while maintaining an acceptable balance between service-user dignity and sample integrity (DH, in press). On the other hand, many opioid users will have a dry mouth on presentation for detoxification and may have genuine difficulty in providing a suitable sample. A further problem of oral fluid testing is that the detection time of drug use is considerably shorter than for urinalysis, generally providing information on use within the last 24 hours (DH, in press; Verstraete, 2004). Drug concentration can also differ depending on the collection method. Stimulation of saliva flow is often used. This can be problematic because the pH for stimulated flow is approximately 8, compared with the basal saliva pH of 6.5. Therefore any drug with a pKa around these values will be substantially affected and may lead to decreased drug concentration (NACB, 2006). Thirdly, there is a lack of evidence on interferences, oral drug residues, and other issues of manipulation that may affect the validity of this matrix (NACB, 2006).
There is limited evidence for the sensitivity and specificity of oral fluid testing products (DH, in press). In a small study (N =15), results obtained by law enforcement officers correlated well with laboratory results for cocaine and amphetamines but were unsatisfactory for detecting heroin and cannabis use (Samyn & van Haeren, 2000). Gronholm and Lillsunde (2001) also found poor sensitivity for detecting benzodiazepines and cannabinoids for oral fluid testing.
5.3.3. Confirmation of screening tests
Confirmatory tests are often needed to reduce false positive results; this may relate to adulteration of the sample or a false interpretation when medications that are chemically similar to the drug of interest are taken legitimately. Conversely, a negative test may not rule out dependence. This may be due to a number of factors, such as the sample being taken some time after drug ingestion, adulteration of the sample or threshold of sensitivity of the analytical procedure in the laboratory.
Confirmation of screening test results is a sophisticated laboratory exercise that requires a considerable investment in skilled staff and dedicated equipment. In general, it is not a service that can be set up or completed rapidly with non-specialised staff or equipment.
The majority of the cases presenting for detoxification will involve opioids detectable by near-patient testing. However, some opioids, including buprenorphine, fentanyl, oxycodone, pethidine and others, are not detectable under standard immunochemical tests and would produce a false negative near-patient test result. If there is uncertainty after a clinical assessment about the drug use or dependence of a service user, confirmatory laboratory testing should be considered.
Confirmatory laboratory testing should be capable of detecting service users who deliberately contaminate their urine with heroin or methadone in order to produce a false positive result. Heroin use may be ascertained in the laboratory by the demonstration of compounds such as 6-monoacetylmorphine, codeine, acetylcodeine, meconin and possible others in urine. There is also a need to confirm the presence of both methadone and its principal metabolite in urine.
The standard of testing in a laboratory providing screening and/or confirmatory services should be high, with appropriately trained staff who all participate in programmes of continuing professional education. There should be appropriate established standard operating and safety procedures in place, and participation in quality assurance schemes that assess not just the analytical capabilities of the laboratory but also the ability of the laboratory staff to interpret results.
In order for a laboratory to react appropriately to an analytical request, the sample must be unequivocally identified and appropriate clinical information must be provided. The format of the report should be clear and should be accompanied by sufficient information to enable the report to be interpreted by the person responsible for the management of the service user’s care. For example, if a report indicates the presence of 6-monoacetylmorphine, then the significance of this should be explained in text below the analytical result; that is, that this metabolite is unique to heroin and can distinguish between the use of codeine prescriptions or poppy seed consumption (which may result in a morphine positive urine sample) and heroin use (Mule & Casella, 1988). The nature of the substance identified should be described accurately and unambiguously; for instance, it would be inappropriate for a near-patient testing instrument that identifies the presence of opioids to report a sample as being positive for heroin.
Where the laboratory is remote from the treatment facility, arrangements must be in place for the rapid and secure electronic reporting of results. Both the laboratory and the care providers should have protocols in place to ensure that results are reported rapidly by the laboratory and reviewed quickly and efficiently by the care providers.
5.3.4. Summary
Testing of biological fluids for misused drugs is an important tool to ensure safety in the care of service users undergoing opioid detoxification. At present, most data on testing is for urinalysis and this remains the most reliable tool for clinical practice. Screening of biological fluids for the presence of opioid drugs should be carried out by techniques that are fit for purpose by adequately trained staff who continue to maintain their skills. Ease of collection, training implications and the equipment required also need to be taken into consideration.
However, the interpretation of tests for the presence of drugs and their metabolites cannot be divorced from knowledge of the clinical circumstances and the donation of the sample. The clinician must also have knowledge of the characteristics of the tests, their limitations and the interpretation of a variety of tests in different settings. If there is uncertainty about the service user’s drug dependence, the clinician may wish to defer initiation of detoxification until confirmatory tests are available. If initiating with only screening tests, the clinician must be certain of clinical dependence or organise detoxification in a setting with adequate observation and dose titration.
Training is important for all clinicians, who should have the support of appropriate and trained laboratory staff. Protocols should be available regarding the practical aspects of taking tests, their refrigeration if appropriate, the need for supervised samples, the extent to which service users should be supervised while providing a sample (that is, the frequency and intrusiveness of the supervision), the need for confirmatory testing and ensuring clinical governance and quality assurance of this aspect of care.
Urinalysis is the most reliable tool for identifying drug use and has higher sensitivity and specificity than oral fluid testing for a number of substances (DH, in press). In addition, urinalysis is substantially less costly than oral fluid testing. Therefore, the routine use of urinalysis is more cost effective, since it represents a more efficient use of limited NHS resources. Healthcare professionals should normally consider using urinalysis for drug testing as the first choice, and consider oral fluid testing only in circumstances were urinalysis is impractical or unacceptable to the service user.
5.3.5. Clinical practice recommendations
- 5.3.5.1.
If opioid dependence or tolerance is uncertain, healthcare professionals should, in addition to near-patient testing, use confirmatory laboratory tests. This is particularly important when:
a young person first presents for opioid detoxification
a near-patient test result is inconsistent with clinical assessment
complex patterns of drug misuse are suspected.
- 5.3.5.2.
Near-patient and confirmatory testing should be conducted by appropriately trained healthcare professionals in accordance with established standard operating and safety procedures.
- 5.3.5.3.
Healthcare professionals should be aware that medications used in opioid detoxification are open to risks of misuse and diversion in all settings (including prisons), and should consider:
monitoring of medication concordance
methods of limiting the risk of diversion where necessary, including supervised consumption.
5.4. PSYCHOMETRIC ASSESSMENT TOOLS
5.4.1. Introduction
The importance of a clinical assessment of opioid (and other drug or alcohol) dependence and monitoring withdrawal before and during detoxification has been discussed above (see Section 5.2). This section is concerned with the use of psychometric instruments as adjuncts to clinical assessment and monitoring.
Crome and colleagues (2006) argue that there are a number of advantages for the use of assessment tools. Recording is standardised, and a checklist of domains ensures that important issues are covered and that multidisciplinary professionals have a common understanding of what has been assessed. Furthermore, the implementation of tools over time can be utilised to demonstrate progress to the service user and to measure outcome. Finally, the use of assessment tools is empirically testable and therefore it is possible to evaluate the reliability and validity of these tools. The reliability and validity of the psychometric tools used to assess dependence and monitor withdrawal are discussed below.
5.4.2. Assessment of dependence
Identification (simple assessment) tools have most recently been reviewed by NICE (2007; National Collaborating Centre for Mental Health, 2008). The present review will focus on assessment of dependence.
There have been a number of recent reviews evaluating assessment tools for drug misuse (Crome et al., 2006; Scottish Executive, 2003; Sperling et al., 2003). Crome and colleagues (2006) and the Scottish Executive (2003) briefly evaluated the assessment tools. Sperling and colleagues (2003) conducted a more detailed consensus-based evaluation of these measures on training/costs, administration, UK relevance, psychometric properties and content, providing an overall summary percentage score of the extent to which these criteria were judged to be fulfilled.
Self-report questionnaires
The Leeds Dependence Questionnaire (LDQ; Raistrick et al., 1994) is a ten-item self-report scale designed to measure dependence on a variety of substances, to be sensitive to change over time (although follow-up data in validation was not long enough to assess this) and to account for the range of mild to severe dependence. Concurrent validity was assessed by comparing the LDQ with the Severity of Opiate Dependence Questionnaire (SODQ) for opioid users and a moderate association was found (r = 0.30). Additionally, there was a high level of internal consistency (Cronbach α = 0.94). Sperling and colleagues’ (2003) consensus-based evaluation of this measure rated it very highly (97%).
The Severity of Dependence Scale (SDS; Gossop et al., 1995) is a short (five-item) self-report scale designed to measure the degree of dependence on a variety of drugs. The SDS is related to behavioural patterns of drug taking such as heroin dose (r = 0.24), frequency of heroin use (r = 0.43) and duration of use (r = 0.27). In addition, it has good concurrent validity, with treatment-seeking participants reporting higher mean scores (t = 10.00, p < 0.001) than non-treatment seeking controls (Gossop et al., 1995). The scale was also found to have a high level of internal consistency (Cronbach α ranging from 0.84 to 0.90 in heroin-user samples). There are mixed reviews of the utility of this measure for clinical practice. Sperling and colleagues (2003), on the same criteria listed above (training/costs, administration, UK relevance, psychometric properties and content), rated this measure the most highly (99%) of all the assessment scales they reviewed. However, another reviewer expressed major concerns about the use of this scale as a measure of dependence due to the lack of items on tolerance and withdrawal (Scottish Executive, 2003).
Clinician-administered questionnaires
The Addiction Severity Index (ASI; McLellan et al., 1980) is a clinician-administered multi-dimensional 200-item measure with seven main areas: medical, employment/support, alcohol, drug, legal, family/social and psychiatric. This assessment tool has been investigated extensively. Makela (2004), in a review of 37 studies on the psychometric properties of the ASI, concluded that there were inconsistent findings on inter-rater reliability, test-retest reliability, and internal consistency for this scale. Furthermore, this scale was not rated very highly (69%) in a review of assessment scales, mainly due to difficulties administering such a large measure in clinical practice, training costs and relevance to the UK (Sperling et al., 2003).
The Opiate Treatment Index (OTI; Darke et al., 1992) is a clinician-administered multi-dimensional measure with sub-scales on drug use, HIV risk behaviour, social functioning, criminality, health and psychological adjustment. Test–retest reliability correlations were large and ranged from 0.88 to 0.96. Associations between the OTI and the ASI generally ranged from r = 0.43 to r = 0.70; however, the correlation between the criminality subscale and the legal subscale of the ASI was very low (r = 0.02). Additionally, agreement between self-report and collateral report (partner or family member) was relatively high. Sperling and colleagues (2003) did not rate this measure particularly highly (73%), citing problems with relevance to the UK and difficulties with administration in clinical practice.
The Maudsley Addiction Profile (MAP; Marsden et al., 1998) is a clinician-administered 60-item scale covering the following domains: substance use, health risk, physical/psychological health and personal/social functioning. Concurrent validity was acceptable, with high correlations (r = 0.72) between the physical and psychological health measure and items adapted from the ASI. Similarly, for the relationship conflict measures of the MAP there were high correlations (r = 0.74) with subscales from the Life Stress and Social Resources Inventory (LISRES). In addition, there was high test–retest reliability averaging 0.94 overall and 0.88 for reported substance use. This measure was also rated highly (96%) by Sperling and colleagues (2003). However, the reviews of both Sperling and colleagues (2003) and the Scottish Executive (2003) advised caution concerning the length of the scale and therefore the ease of administration in clinical practice. As a response to such criticisms, the MAP has recently been adapted into a shorter (20-item) self-completion version (Luty et al., 2006). There were relatively large correlations (r = 0.70) between the adapted self-completion and the original interviewer-completion version of the MAP.
The Christo Inventory for Substance-Misuse Services (CISS; Christo et al., 2000) is a ten-item clinician-administered measure including social functioning, general health, sexual/injecting risk behaviour, psychological functioning, occupation, criminal involvement, drug/alcohol use, ongoing support, compliance and working relationships. Relatively large correlations were found with the OTI (generally ranging from r = 0.70 to 0.91). There was also good inter-rater reliability with Pearson’s correlations of r = 0.84 and an intraclass correlation of 0.82 (Christo et al., 2000). The reviews of both Sperling and colleagues (2003) and the Scottish Executive (2003) suggested problems with the content of this measure, suggesting it may be too simplistic.
5.4.3. Monitoring of withdrawal
The most important aspects of monitoring objective and subjective withdrawal symptoms in clinical practice are to determine that over- or under-prescribing is not occurring and that the service user is comfortable on his or her dose. This is primarily monitored by clinical assessment, but the use of psychometric measures can aid this process.
Scales measuring withdrawal are commonly categorised as objective (clinician-rated) or subjective (self-report). There are several scales that have been developed to monitor the withdrawal process; these include: the Clinical Opiate Withdrawal Scale (COWS; Wesson & Ling, 2003), Opiate Withdrawal Scale (OWS; Bradley et al., 1987), Short Opiate Withdrawal Scale (Gossop, 1990) and the Subjective and Objective Withdrawal Scales (Handelsman et al., 1987).
The self-reported OWS was assessed during a 20-day detoxification trial of 84 participants (Bradley et al., 1987). The pattern of withdrawal as measured by the scale was as expected. As methadone dose was reduced, a rise in distress was reported that faded by the end of the third week to a total withdrawal score in the normal range (derived from a non-dependent control group). There was a relatively small correlation (r = 0.25) between the self-report OWS and nurse observation of withdrawal, although correlations between nurse observation and the OWS were much higher when the nurse-observed rating was high (r = 0.71). Gossop (1990) compared the Short Opiate Withdrawal Scale (10 items) with the OWS (32 items). A very high correlation (r = 0.97) was found between these measures, suggesting the usefulness of the shorter version.
The Subjective and Objective Opiate Withdrawal Scales were assessed for 32 participants admitted for inpatient detoxification (Handelsman et al., 1987). Significant changes were found for both scales at the stabilisation stage of the trial and after a naloxone challenge.
The COWS is a clinician-rated measure. There appears to be little validation of this measure, with the exception that all items have been validated in previous measures (Wesson & Ling, 2003).
5.4.4. Summary
The development of psychometric tools to assess dependence and monitor withdrawal is still at an early stage. Although data were relatively sparse for most measures, some had reasonable reliability and validity. The use of reliable and valid assessment tools may aid the process of conducting a clinical assessment and monitoring withdrawal during the process of detoxification.
