Systematic review

Substance‐related disorders Pharmacology
Volume 208 - Issue 10

Identifying and treating codeine dependence: a systematic review

Authors:  Suzanne Nielsen, Tim MacDonald and Jacinta L Johnson

Med J Aust 2018; 208 (10): 451-461. || doi: 10.5694/mja17.00749
Published online: 12 February 2018
Download PDF (367 KB)
Share: Share on Blue Sky Share on Facebook Share on LinkedIn Share by email

Abstract

Objectives: Codeine dependence is a significant public health problem, motivating the recent rescheduling of codeine in Australia (1 February 2018). To provide information for informing clinical responses, we undertook a systematic review of what is known about identifying and treating codeine dependence.

Study design: Articles published in English that described people who were codeine-dependent or a clinical approach to treating people who were codeine-dependent, without restriction on year of publication, were reviewed. Articles not including empirical data were excluded. One researcher screened each abstract; two researchers independently reviewed full text articles. Study quality was assessed, and data were extracted with standardised tools.

Data sources: MEDLINE and EMBASE were searched for relevant publications on 22 November 2016. The reference lists of eligible studies were searched to identify further relevant publications. 2150 articles were initially identified, of which 41 were eligible for inclusion in our analysis.

Data synthesis: Studies consistently reported specific characteristics associated with codeine dependence, including mental health comorbidity and escalation of codeine use attributed to psychiatric problems. Case reports and series described codeine dependence masked by complications associated with overusing simple analgesics and delayed detection. Ten studies described the treatment of codeine dependence. Three reports identified a role for behavioural therapy; the efficacy of CYP inhibitors in a small open label trial was not confirmed in a randomised controlled trial; four case series/chart reviews described opioid agonist therapy and medicated inpatient withdrawal; two qualitative studies identified barriers related to perceptions of codeine-dependent people and treatment providers, and confirmed positive perceptions and treatment outcomes achieved with opioid agonist treatments.

Conclusion: Strategies for identifying problematic codeine use are needed. Identifying codeine dependence in clinical settings is often delayed, contributing to serious morbidity. Commonly described approaches for managing codeine dependence include opioid taper, opioid agonist treatment, and psychological therapies. These approaches are consistent with published evidence for pharmaceutical opioid dependence treatment and with broader frameworks for treating opioid dependence.

PROSPERO registration: CRD42016052129.

Codeine is globally the most frequently used opiate,1 and its consumption is increasing. In Australia, 27 780 234 packs of codeine-containing analgesics were supplied by community pharmacies during 2013, a rate of 1.24 packs per person.2 In New Zealand, most of Canada, South Africa, Ireland, and the United Kingdom, codeine is available over the counter, usually combined with simple analgesics such as paracetamol or ibuprofen;3 until recently, it was also available without prescription in Australia and France. Products containing greater amounts of codeine are generally only available on prescription.3

Codeine has low affinity for and intrinsic activity at µ-opioid receptors, and is considered a prodrug; its analgesic effects depend largely on its being converted to morphine by the polymorphic cytochrome P450 isoenzyme (CYP) 2D6.4,5 Genetic variability in the activity of CYP2D6 underlie interperson differences in the analgesia achieved and the risk of opioid toxicity.6 Tolerance can develop after a relatively short period of regular use.7-9

In view of the limited evidence that adding low dose codeine (< 30 mg) to simple analgesics increases pain relief,10-15 the variability in its metabolism, and the availability of opioids with more predictable effects, the role of codeine in pain management is contentious.16,17

The liability of codeine to be misused has been shown in a randomised, double blind, placebo-controlled drug administration study,18 and has been documented in several case series.19,20 Although the prevalence of codeine dependence is unknown, the harms associated with overuse are well established, including serious morbidity causing great cost to the health care system.21

Some harms associated with codeine overuse are directly related to prolonged intake, but many serious consequences stem from concomitant overconsumption of ibuprofen or paracetamol in combination products.19 Sequelae of supratherapeutic ibuprofen ingestion secondary to codeine dependence that require intensive care have been described, including several codeine-related deaths.22 As a result, access to over-the-counter codeine has been restricted or removed in Manitoba (February 2016), France (July 2017), and Australia (February 2018).23-25

In order to respond appropriately, we need to identify people who are codeine-dependent. There has been greater awareness of dependence with the imminent rescheduling of codeine in Australia. Both in Australia and internationally, presentations to addiction treatment services have increased,26-28 but treatment approaches for codeine dependence are poorly defined. The purpose of our systematic review was to identify the characteristics of people who are codeine-dependent, and to define approaches for identifying codeine dependence.

Methods

Search strategy

We searched MEDLINE and EMBASE on 22 November 2016 for the following terms: “codeine”, “dependence”, “substance-related disorders”, “opioid-related disorders”, “behaviour, addictive”, and “substance withdrawal syndrome” (online Appendix, table 1). We restricted our search to human studies published in English; there was no restriction on year of publication. The reference lists of eligible studies were searched to identify further relevant publications.

One reviewer (SN, JJ or TM) independently examined the titles and abstracts of identified articles. The full text of relevant articles was independently assessed for inclusion by two authors, and reasons for exclusion documented as appropriate. Inter-reviewer disagreement about inclusion was resolved by consensus among all three authors.

Study inclusion criteria

We included studies that described people who were codeine-dependent (identification studies) or any clinical approach for treating people who were codeine-dependent (treatment studies).

Data extracted from identification studies included study characteristics (author, location, design, quality rating) and population characteristics (participant age, sex, employment, mental health, pain and substance use history, adverse effects related to codeine use, and management of adverse effects).

Treatment studies included randomised and non-randomised controlled trials, quasi-experimental, before-and-after studies, prospective and retrospective cohort studies, case–control studies, analytic cross-sectional studies, qualitative studies, and case reports and series. Treatment outcome measures included change in codeine use (days of use or amount used), retention in treatment, adverse events and other outcomes related to codeine use, opioid dependence, and pain.

Exclusion criteria

Reports limited to describing the clinical applications or pharmacology of codeine or other opioids, reports that did not separately report codeine-related data, and articles without empirical data (eg, letters, commentaries, reviews) were excluded (online Appendix, table 2).

Assessment of methodological quality (identification studies)

The quality of descriptive studies was assessed with a modified version of the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Analytical Cross Sectional Studies.29 To enable application of a single tool to all study methodologies, the JBI tool was reduced from eight to five items, and an item from the Evidence-Based Librarianship (EBL) Critical Appraisal Tool for assessing sample bias was added30 (online Appendix, table 3). The range for total scores was 0–6, higher scores reflecting higher quality.

Grading of evidence (treatment studies)

Studies examining treatment approaches were scored for quality according to GRADE criteria.31

Data collection

Data were extracted with a standardised data extraction tool into an Excel (Microsoft) spreadsheet. The tool was piloted and reviewed before being finalised.

Data synthesis

Findings were qualitatively and quantitatively synthesised when population characteristics were reported in a manner that enabled this approach. Meta-analysis of treatment studies was not possible because of the heterogeneity of study designs. When individual patient data were reported, details were extracted at the patient level to enable synthesis of patient characteristics.

Results

Of the 2150 articles initially identified, 41 were eligible for inclusion in our analysis (Box 1). The mean study quality score of the included articles was 3.0 (standard deviation [SD], 1.1).

Identifying codeine dependence

Fourteen reports described samples of patients who were codeine-dependent (Box 2; online Appendix, table 4A); 22 described presentations by individual patients (Box 3, Box 4; online Appendix, tables 4B and C). No studies reported developing an approach for identifying people with codeine dependence as an aim, but two reported applying the Severity of Dependence Scale (SDS)32 for defining codeine dependence (cut-off score, 5).33,34

Analyses of administrative data

Three studies examined data from administrative sources on the treatment of people for opioid dependence.28,35,36 An Australian study compared codeine-related treatment episodes with those for patients for whom another prescribed opioid or heroin was the chief drug of concern. The proportion of women among those treated for codeine dependence declined from 70% in 2002 to 47% in 2011; people for whom codeine was the drug of concern were on average older and less likely to have a history of intravenous and illicit substance use than those treated for misuse of stronger prescription opioids or heroin.28 A study of codeine prescriptions in Norway found that 0.5% of all codeine recipients in 2005 were likely to be using codeine problematically (annual prescription level exceeding twice the maximum daily dose for 12 months); further, high dose prescribing of benzodiazepines was more prevalent in this group.35 A South African study of national data on treatment for substance misuse (alcohol, pharmaceutical and illicit drugs) found that 2.5% of admissions involved codeine, and that codeine was recorded as the primary substance of concern for 0.8% of patients.36

Quantitative convenience samples

Two studies included convenience samples of people who reported using codeine. In an Australian web-based survey, 137 codeine-dependent participants were compared with 633 non-dependent participants; characteristics associated with dependence (assessed with the SDS) included taking higher than recommended doses, experiencing psychological distress, previous drug treatment, and chronic pain.33 A 1999 Canadian postal survey on prescribed and over-the-counter codeine (participants recruited via newspaper advertisements) found that 37% of respondents met DSM-IV criteria for codeine dependence, most of whom reported chronic pain and family histories of substance use problems.20

Case–control study

A prospective case–control study described patients attending an addiction medicine clinic in China who were dependent on a codeine-containing cough syrup.37 This imaging study found that the patients, who exhibited increased impulsivity, had cortical white matter microstructural abnormalities.37

Qualitative studies

Three qualitative studies have examined the perceptions of pharmacists and codeine-dependent people.34,38,39 A British author38 described the perception that dependence is not identified early, the challenges posed by the stigma attached to dependence, the fact that codeine-dependent people saw themselves as different to users of illicit opioids, and medical reasons for initiating codeine use.38 A recent Irish study similarly described social stigma as a treatment barrier, and reported emotional distress as a driver for codeine use.34

Two typologies of codeine dependence have been proposed:38,39 users who do not exceed therapeutic doses, and codeine-dependent people who consume high doses of the drug. The second group is characterised by severe dependence and harms. Additional typologies include recreational users39 and people who slightly exceed recommended doses.38

Retrospective chart reviews and case series

Five case series or retrospective chart reviews19,40-43 identified common features of people with codeine dependence, including higher proportions of women than among those treated for misuse of other opioids, histories of problematic alcohol use, mental health comorbidity, and serious side effects resulting from using combination medicines containing codeine, including one death.41 Prior heroin dependence was rare.

Reports on individual patients

Twenty-two reports described 49 individuals who were codeine-dependent44-65 (Box 3, Box 4; online Appendix, tables 4B and 4C). Twenty-three were women (mean age, 42 years [SD, 9 years]). Of the 15 people for whom data on employment status were reported, nine were employed. Acute harms and information on treatment approaches were described. Inherent to cases of acute harm were complications attributable to the co-medications paracetamol and ibuprofen, including distal renal tubular acidosis, hypokalaemia, gastritis and other enteropathies, medication overuse headache, hepatic necrosis, hypoalbuminaemia, microcytic anaemia, and weight loss. In a case series of 27 patients,19 most had initiated codeine to treat pain but later escalated their intake for other reasons; this was also reported in many case reports.

Acute management of harms was characterised by inpatient hospital management of serious problems (often requiring intensive care) to restore electrolyte balance,47,58,59 manage gastrointestinal symptoms44,53 (including bowel resection),56 and to assess or treat hepatotoxicity.50,64 Opioid withdrawal was managed with symptomatic medications and buprenorphine; potential hepatotoxicity was managed with acetylcysteine.64

Psychiatric comorbidity in people with codeine dependence was dominated by high prevalence conditions (depression and anxiety disorders).20,27,44,48,53,58,60,61 Some reports described prior or comorbid addictions (benzodiazepines, opioids, alcohol),34,35,40,42,60 and mental health conditions.20,40-42,60 Bipolar disorder,27,42 obsessive–compulsive disorder,62 relationship breakdown,47 suicide behaviour,51,56 and bereavement or loss47 were also described.

Individual patient reports described treatment approaches, including attempted self-management,45,65 use of symptomatic medications,60 prescribed codeine or dihydrocodeine49,52 or buprenorphine and naloxone,52,54,65 and detoxification with methadone.48 Management of mental health symptoms with antidepressants and behavioural therapies was described.48,54,55,65 One notable case combined methadone taper, cognitive behavioural therapy, and relaxation strategies, enabling codeine cessation and self-management of pain.48 Resolution of presenting complaints (obsessive–compulsive disorder, medication overuse headache) after codeine cessation was described.62,63

Treatment studies

Ten studies described treatment approaches in detail (Box 5; online Appendix, table 4D). Common medication-based approaches included taper from codeine with symptomatic medications such as clonidine or benzodiazepines,42,67 buprenorphine maintenance,19,27 CYP inhibitors,66,68 and gradual self-managed taper.38 Positive outcomes for opioid agonist treatment (methadone and buprenorphine with or without naloxone) were described.19,27,34,38,40,43,69 The role of internet support groups38 and psychosocial treatments, including cognitive behavioural therapy, were highlighted in some studies.42,67,69

Two studies tested the hypothesis that preventing the O-demethylation of codeine to morphine with CYP inhibitors would reduce codeine use.66,68 Initial promising results from an open label pilot study of fluoxetine (14 subjects)66 were not replicated in a small randomised controlled trial that compared the effectiveness of fluoxetine or quinidine (two potent CYP2D6 inhibitors) with placebo.68

A retrospective review of inpatient admissions described taper with clonidine and benzodiazepines, combined with an intensive 4-week mental health treatment program.42 Patients had a mean stay of 42 days (SD, 23 days), with withdrawal symptoms requiring treatment for a mean 16 days (SD, 10 days). Taper with buprenorphine was described by an Australian study which noted that codeine dependence was more likely to be treated with taper rather than maintenance.27 Relapse after taper was not uncommon.34,42

A single arm study (11 patients) found that cognitive behavioural therapy could significantly reduce codeine use — six patients ceased using codeine altogether — and neuropsychological functioning was improved by codeine reduction or cessation without deterioration in pain or quality of life.67 A mixed methods study also highlighted the role of psychological therapies.69

Seven studies described treatment with opioid agonists,19,27,34,38,40,43,69 four of which did not report outcomes.19,34,40,69 One small retrospective cohort study described high retention rates for patients treated with buprenorphine over 28 days (median daily dose, 12–16 mg); one patient described initial sedation that necessitated reducing the dose.43

Two qualitative studies described positive experiences and outcomes for treatment with methadone and buprenorphine, despite patient concerns about the treatment experience and the clinic environment.34,38

According to GRADE criteria, the quality of evidence from treatment studies was very low to low; most studies were retrospective and descriptive, and all had small sample sizes.

Discussion

Our review of codeine-dependent people indicates that approximately equal proportions of men and women are involved; their mean age is greater than for patients treated for problematic use of other opioids, the prevalence of mental health comorbidity is high, identification of dependence is often delayed, and patients experience serious complications associated with excessive consumption of combination products that include codeine. Problematic codeine use was associated with mental health problems. The quality and methodology of the studies we assessed varied considerably, but their depictions of the features associated with codeine dependence were consistent, describing a clinically challenging area in which under-reporting is highly likely. The reports highlight the importance of asking about the use of non-prescribed analgesics in a range of health care situations, particularly when gastrointestinal complications are identified. The diversity of those affected and the high level of morbidity suggest that population level interventions are required for screening and prevention wherever codeine is available over the counter. Careful questioning about recent patterns of use, the reasons for taking codeine, and withdrawal symptoms upon cessation may help identify when a patient should be comprehensively assessed for an opioid use disorder.

Treatment approaches include self-management with internet support, psychological treatments, symptomatic medications for opioid withdrawal, and opioid agonist treatments. In particular, buprenorphine treatment undertaken according to current guidelines was commonly described. Studies of opioid taper found that relapse was common (consistent with taper for opioid dependence in general). Taken together, the treatment studies and case reports provide evidence that opioid agonist treatments, combined with psychosocial adjuncts, may be suitable and acceptable to patients. The evidence, albeit low in quality, indicates that positive treatment outcomes could be achieved with these approaches.

In the absence of specific high quality evidence, judgements about approaches for treating people with codeine dependence must be based largely on studies of opioid dependence. The effectiveness of treatment with methadone and buprenorphine has been reported, and maintenance is more effective than withdrawal and detoxification for people who are dependent on pharmaceutical opioids70 or opioids in general.71 Research on selecting patients for treatment with opioid agonists is limited. According to the general principles of treatment, diagnosis of opioid dependence must first be confirmed.72 A stepped care approach with less intensive treatment (eg, taper, counselling) for low severity dependence is recommended by national guidelines.72 Patients who unsuccessfully attempt taper may be considered for maintenance opioid agonist treatment, which achieves better treatment outcomes than detoxification for pharmaceutical opioid dependence.71 Because of wide variations in codeine metabolism, predicting opioid requirements with dose conversion tables is challenging;43 for this reason their use is discouraged.

Psychological adjunct therapies can be beneficial,73 but the role of psychosocial interventions as accompaniments to opioid agonist treatments requires further research.74 The high prevalence of mental health comorbidities and the preference of patients for online support may indicate that online interventions for managing comorbidity may be useful. In general, the role of pharmacological treatments for depression or anxiety at the start of treatment is unclear. It is recommended that comorbidities are assessed after a period of abstinence because of the potential for diagnostic uncertainty caused by the acute effects of opioid toxicity and withdrawal.75

The treatment setting is also important. People consuming larger amounts of opioids together with sedatives (eg, benzodiazepines) are a population at greater risk, and referral to a specialist may be required.72 Characteristics that may indicate that patients are appropriate for managing in primary care include being employed, having social support, and not having another substance use disorder or a history of illicit drug use.

Medication overuse headache

Headache is a common reason for initiating codeine use by patients who develop dependence.19,61 Paradoxically, medication overuse headache — in this context, exacerbation of a pre-existing headache disorder by excessive intake of codeine — is another potential complication of codeine dependence.48,55,63 Data that might guide the management of codeine overuse headache specifically have not been published. Management of opioid-related medication overuse headache usually consists of patient education, opioid withdrawal, and the initiation of prophylactic agents,76-78 often in an inpatient setting.76 Medication overuse headache that results from overusing analgesics, compared with overuse of triptans, is associated with a greater withdrawal headache duration (about 10 days),79 with meaningful improvement only after 12 weeks or more,80 and high relapse rates (eg, 71% at 4 years81).

Limitations of our analysis

Comparing codeine dependence in different groups of patients was made difficult by changing usage patterns over time, subgroup heterogeneity, and probable under-reporting of codeine use. Methodological constraints included low participant numbers, selection biases (admissions, help-seeking or co-medication sequelae as a proxy for neuroadaptation to codeine), and a lack of objective and standardised criteria for determining codeine dependence. Many studies employed internet-based recruitment or data collection,33,38 potentially limiting the generalisability of findings to users without regular internet access, but this might be offset by the ability to reach users who are otherwise difficult to reach. Some studies did not specify whether codeine was prescribed or obtained over the counter, but most reports were concerned with over-the-counter codeine. Many studies that included codeine-dependent people were excluded from our analysis because they did not separately describe codeine dependence; this particularly applied to studies of medication overuse headache. Nevertheless, our review is the most comprehensive synthesis of data on the phenomena of codeine dependence, and we have described a range of potential treatment responses, including medication- and non-medication-based treatments.

Conclusion

Codeine dependence can be identified by screening patients who present with acute complications associated with taking combination analgesics, and by routine questioning about over-the-counter medication use. Common treatment approaches include detoxification and opioid agonist treatment. Clinical leadership in providing guidance about how to identify and treat individuals with codeine dependence is required as a matter of public health.

Box 1 – Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) diagram of study selection

Box 2 – Studies describing characteristics of people who were codeine-dependent

Study

Design, location

Sample size, type

Sex (women)

Age (years)

Quality rating

Sproule 199920

Survey, Canada

339 people who had used codeine 3 days/week for at least 6 months

51%

mean, 43.5 (range, 18–82)

5

Fredhein 200935

National prescription database, Norway

385 190 people prescribed codeine during 2005

56%

mean, 52.3 (SD, 18.8)

5

Frei 201019

Case series of patients who presented or were referred to hospital addiction medicine, Australia

27 treatment admissions for serious harms from over-the-counter ibuprofen–codeine products

48%

= 20

3

Thekiso 201042

Retrospective chart review, Ireland*

20 inpatients admitted with over-the-counter codeine abuse or dependence

65%

mean, 49 (SD, 23)

4

McAvoy 201140

Case series, New Zealand*

15 patients presenting for over-the-counter codeine detoxification

46%

mean, 44 (range, 30–60)

3

McDonough 201141

Retrospective chart review, Victoria

32 people referred to hospital drug and alcohol services for excessive over-the-counter codeine use (Sept 2005 – Sept 2010)

72%

median, 38

2

Nielsen 201133

Online survey, Australia

137 people who met criteria for codeine dependence (subset of 800 people reporting over-the-counter codeine use)

66%

mean, 37 (SD, 13)

3

Cooper 201338

Qualitative interviews, United Kingdom*

25 participants recruited from internet support groups

52%

20–60s

3

Nielsen 201339

Survey (qualitative research methodology), Australia

20 codeine-dependent people

60%

mean, 39 (SD, 11)

2

Dada 201536

National administrative treatment dataset, South Africa

425 codeine-related admissions to alcohol and other drug treatment centres, 137 with codeine as drug of primary concern

25%

11–70 (20–39 years: 59%)

2

Nielsen 201528

Cross-sectional study of national administrative datasets, Australia

4424 codeine treatment episodes

70% (2002) 47% (2011)

mean, 36

2

Nielsen 201527

Retrospective chart review, Australia*

135 (53 codeine-dependent)

66%

mean, 38 (SD, 8)

4

Qiu 201537

Case–control study mapping brain differences, China

60 (30 codeine-dependent people, 30 controls)

7%

mean, 25

2

Van Hout 201534

Qualitative interviews, Ireland*

21 participants recruited through drug treatment centres (in treatment or in recovery for over-the-counter codeine use)

57%

mean, 39 (range, 26–62)

2

SD = standard deviation. * Study also included in (treatment approaches). For further details, see online Appendix, table 4A.

Box 3 – Individual patient reports: acute description or management of codeine-related harms

Study

Location of patient (age, sex)

Harms from codeine use

Details of codeine dependence

Treatment

Faierman 197350

USA (male, 30)

Hepatic injury with extensive fibrosis attributed to terpin hydrate component of cough syrup, as opposed to codeine

16–20 ounces codeine cough syrup daily

Not reported

Wylie 199464

Scotland (female, 37)

Elevated liver enzyme levels

Up to 30 × 500 mg paracetamol/30 mg codeine phosphate per day

Treated for potential hepatotoxicity with acetylcysteine; discharged after 2 days

Dyer 200447

United Kingdom (male, 49)

Perforated duodenal ulcer 3 years previously (inappropriate ibuprofen use), hypokalaemia

30 × 200 mg ibuprofen/12.8 mg codeine in the 3 days before admission. GP reported their taking 24 tablets at once previously

Serum potassium corrected with intravenous potassium therapy; advised that Nurofen Plus misuse caused his recurrent hypokalaemic episodes; offered assistance

Dutch 200846

Australia (female, 39)

Anterior gastric antrum ulcer and 2.6 L of green turbid fluid in the peritoneal cavity

16–24 × 200 mg ibuprofen/12.8 mg codeine per day for 3 weeks

Transferred to intensive care in another hospital

 

Australia (male, 41)

Gastric antrum ulcer with gross abdominal contamination

“A packet” of 200 mg ibuprofen/12.8 mg codeine per day for one year

Postoperatively, patient offered inpatient drug treatment, absconded before transfer arranged

Karamatic 201153

Australia (male, 42)

Multiple jejunal ulcers with early structuring consistent with NSAID enteropathy

10 × 200 mg ibuprofen/12.8 mg codeine phosphate per day

40 mg omeprazole, 15 mg mirtazapine daily; 5 mg oxycodone every 4–6 h as needed

 

Australia (female, 41)

Multiple web-like strictures with circumferential ulceration throughout small bowel consistent with NSAID enteropathy

20 × 200 mg ibuprofen/12.8 mg codeine phosphate per day, 5 years

Iron supplement

 

Australia (male, 41)

Multiple jejunal ulcers

10–12 × 200 mg ibuprofen/12.8 mg codeine phosphate per day, > 5 years

Omeprazole and amitriptyline daily; hyoscine butylbromide, paracetamol and codeine and tramadol as needed

Ng 201158

Australia (female, 32)

Oesophageal erosions, benign gastric ulcer, enlarged, oedematous kidneys without nephrocalcinosis

20 × 200 mg ibuprofen/12.8 mg codeine phosphate per day

Electrolyte replacement

 

Australia (male, 37)

Progressive muscle weakness, low serum potassium level, biochemical features consistent with distal renal tubular acidosis

24 × 200 mg ibuprofen/12.8 mg codeine per day

Electrolyte replacement, buprenorphine maintenance therapy

 

Australia (female, 45)

Microcytic anaemia, gastric antral ulceration with a peptic oesophageal stricture

50 × 200 mg ibuprofen/12.8 mg codeine phosphate per day

Electrolyte replacement

 

Australia (male, 40)

Generalised weakness associated with hypokalaemia, with distal renal tubular acidosis

1.4–2.0 g ibuprofen in codeine combination product per day for 3 months

Electrolyte replacement

Page 201159

Australia (females, 35, 39; males, 41, 55)

Renal tubular acidosis, normal anion gap metabolic acidosis

Longstanding misuse of ibuprofen (5–18 g/day) and codeine (320–1152 mg/day) in over-the-counter medications

Biochemical recovery of all patients; two patients required intensive care admission for central venous access and potassium replacement

Lake 201356

Australia (male, 35)

Small bowel stricture secondary to NSAID, loss of partner and employment

Up to 90 × 200 mg ibuprofen/ 12.8 mg codeine phosphate per day, clear salience and neuroadaptation

Exploratory laparotomy and small bowel resection; patient controlled concomitant ketamine use; community drug treatment on discharge

Roussin 201361

France (females, 38, 38, 42, 47; males, 42, 55)

Included depressive mood, and constipation and vertigo

120–200 mg codeine phosphate in combination product with paracetamol per day for 1–10 years

Not reported

Ammit 201644

Australia (female, 39)

Gastric erosion and renal tubular acidosis

520 mg/day (over-the-counter ibuprofen combination) for past year; increased with physical and psychological stress

Symptomatic medication (diazepam, paracetamol, baclofen); balloon enteroscopy/dilation (small bowel obstruction); education about harms of ibuprofen. Following admission: opioid substitution therapy, counselling and 12-step program

NSAID = non-steroidal anti-inflammatory drug. For further details, see online Appendix, table 4B.

Box 4 – Individual patient reports: treatment of codeine dependence, with or without management of acute harms

Study

Location (sex, age)

Harms from codeine use

Details of codeine dependence

Treatment and outcome

Gruber 194851

USA (male, 57)

Neuroadaptation and decline in functioning, loss of weight, likely maintenance of chronic pain symptoms, suicide following withdrawal

Intravenous codeine up to 4.8 g daily in weeks before hospitalisation; 660 mg per day in previous months

Reducing doses of intravenous codeine over 12 days; 100 mg pethidine iv 4–6 times per day (days 3–11), acetylsalicylic acid injections (days 12–15)
Withdrawal syndrome tolerated with some discomfort; committed suicide after discharge

Vaughan 196763

New Zealand (male, 53)

Renal failure (fatal)

8–12 aspirin/phenacetin/codeine tablets daily, several years

Supportive therapy

New Zealand (female, 39)

Renal failure (fatal)

50 aspirin/phenacetin/codeine tablets per week

Symptomatic treatment for renal failure

New Zealand (female, 70)

Analgesic nephropathy

8 aspirin/phenacetin/codeine tablets per day, 20 years

 

New Zealand (male, 28)

Possible medication overuse headache

6–20 aspirin/phenacetin/codeine tablets per day, 20 years

Education on link between symptoms and analgesic use; patient ceased analgesics

Senjo 198962

Japan (male, 34)

Suspected codeine use contributed to obsessive–compulsive disorder in both patients

10-year history of codeine use

Inpatient stay (2 months); withdrawal symptoms after 5 days codeine-free
Obsessive–compulsive disorder improved after withdrawal

 

Japan (male, 35)

10-year history of codeine use

Patient was violent after 2 days, transferred to another hospital
Returned 2 months later with complete remission of symptoms

Bedi 199145

India (male, 42)

Dependence and opioid withdrawal symptoms

Two bottles Phensedyl (total content: 450 mg codeine, 366 mg ephedrine, 180 mg promethazine) per day

Loperamide, diazepam, nitrazepam; supportive psychotherapy and family counselling advocated; drugs reduced over 10 days

Eng 199648

USA (male, 54)

Medication overuse headache

6–15 × paracetamol/codeine tablets

Detoxification (methadone); referral to anxiety disorders program (diagnosed with GAD), CBT, taught relaxation
Self-managed analgesic use (reduced analgesic use to paracetamol twice a week or less), developed alternative strategies for psychological symptoms

Lake 200855

USA (female, 39)

Transformation of episodic to daily headache

10 butalbital with codeine and acetaminophen tablets per day for pain control for past year

Withdrawn from butalbital, codeine; coached in relaxation techniques. After multiple admissions: weekly psychotherapy, formal substance abuse program, observed urine drug screens). CBT, pain management.12-step program
Ongoing relapse, eventually ceased substance use, diagnosed with fibromyalgia and prescribed opioids

Evans 201049

New Zealand (male, 35)

Acute gastric ulcer, severe gastritis and post-bulbar duodenitis with active bleeding

More than 100 × 200 mg ibuprofen/12.8 mg codeine phosphate per day for back pain

Reducing codeine dose prescribed; counselling
Gastrointestinal symptoms healed, but balloon dilation of pyloric stenosis required

Robinson 201060

New Zealand (male, 53)

Gastric ulcer, gastric bleeding, hepatotoxicity

60–80 × 200 mg ibuprofen/12.8 mg codeine phosphate per day, 2 years

Treatment not reported
Many patients reported significant opioid withdrawal symptoms despite treatment with ancillary medications

 

New Zealand (female, 31)

Peptic ulcer and anaemia

48 × 200 mg ibuprofen/12.8 mg codeine phosphate per day, 2 years

 

 

New Zealand (female, 63)

Gastric ulcer

20 × 200 mg ibuprofen/12.8 mg codeine phosphate (and prescription codeine) per day, 3 years

 

 

New Zealand (female, 47)

“Inflammatory bowel disease”

Up to 72 × 200 mg ibuprofen/12.8 mg codeine phosphate per day, one year

 

 

New Zealand (male, 52)

Ileal resection

Up to 80 × 200 mg ibuprofen/12.8 mg codeine phosphate per day, one year

 

 

New Zealand (female, 31)

Gastric ulcer and bleeding (previous gastrectomy)

Up to 120 × 200 mg ibuprofen/12.8 mg codeine phosphate per day, 2 years

 

 

New Zealand (male, 35)

 

Up to 48 × 200 mg ibuprofen/12.8 mg codeine phosphate per day, 2 years

 

Hard 201452

United Kingdom (female, mid-20s)

Neuroadaptation, exclusion of other activities, financial

10-year history of codeine dependence (initially prescribed)

GP changed from codeine to dihydrocodeine as a harm minimisation strategy (approximately 2940 mg dihydrocodeine daily). Buprenorphine/naloxone (maintenance: 10 mg buprenorphine/2.5 mg naloxone); engaged with recovery support services and psychosocial counselling; 12-step program
Initially mild precipitated withdrawal; stabilised and returned to work

Marr 201557

Scotland (female, 24)

Dependence

Initially self-medication for dental pain, escalated to prescribed and over-the-counter opioids

Stabilised on 16 mg buprenorphine/4 mg naloxone, transferred to mono-product for pregnancy
Transferred to community prescriber, reported stigma and discomfort with drug treatment clinic environment

Kean 201654

United Kingdom (male, mid-30s)

Neuroadaptation, relationship disharmony, rebound headaches, hypoalbuminaemia, ALT levels elevated

Codeine prescribed by GP for back pain, later supplemented with illicit codeine, escalating over years to 250 mg/day

Buprenorphine/naloxone induction (up to 8 mg/2 mg daily), tapered over 4 months. Anxiety/depression at end of taper responded to fluoxetine, counselling
Abstinent, functioning and intact relationships

Van Hout 201665

Ireland (female, 57)

Estranged from family, unable to work, episode of haematemesis

Use escalated from 12 to 24–48 tablets per day over 3 years after fracture

Stabilised on 4 mg buprenorphine/1 mg naloxone, counselling every 2 weeks
Continues treatment in pharmacy setting; plan after 2 years to begin taper

 

Ireland (female, 44)

Identified because of high volume of sick notes (impact on employment)

Escalated use of over-the-counter codeine (about 36 tablets per day) at time of traumatic event

Commenced buprenorphine–naloxone, venlafaxine for depression, propranolol for migraine and omeprazole for a peptic ulcer
Stabilised on a 14 mg buprenorphine/3.5 mg naloxone, migraines largely resolved; soon after treatment, antidepressant treatment ended. Returned to work, planned reduction of buprenorphine

 

Ireland (male, 45)

Perforated ulcer requiring surgical repair, three later ulcers, multiple surgical admissions for epigastric pain and gastrointestinal bleeding

Long history of over-the-counter codeine misuse causing life-threatening morbidity

Several failed detoxifications; attempts to stabilise on maintenance dose of codeine failed. Prescribed buprenorphine/naloxone
Overdosed on benzodiazepines, hospitalised, buprenorphine withdrawn. After restabilisation, maintained on 12 mg buprenorphine/3 mg naloxone

 

Ireland (male, 44)

Not specifically reported

Escalating amounts of over-the-counter codeine–ibuprofen (up to 72 tablets per day) over several years

Buprenorphine–naloxone, psychosocial interventions. Initial relapse (attempt to self-detoxify), recommenced on higher dose
Initially stabilised on maintenance dose of 8 mg buprenorphine/2 mg naloxone daily; recommenced and stabilised on 12 mg buprenorphine/3 mg naloxone daily, ongoing counselling

ALT = alanine transaminase; CBT = cognitive behavioural therapy; iv = intravenous. For further details, see online Appendix, table 4C.

Box 5 – Treatment approaches for codeine dependence

Study

Design, location

Sample size (sex), age

Treatment approach and outcomes

Evidence quality (GRADE)

Romach 200066

Open label pilot trial, Canada

14 (36% women); mean, 41 years (SD, 6.6 years)

20 mg fluoxetine per day as CYP2D6 inhibitor (began to taper opioids over 8 weeks of active treatment)
All patients reduced opiate use (range, by 30–100%). Depressive symptoms also reduced

Low

Fernandes 200268

Double blind, randomised controlled trial, Canada

30 assessed, 17 started treatment (65% female), mean, 40 years (SD, 12 years)

All patients received brief behavioural therapy. Two weeks of baseline monitoring were followed by 8 weeks of daily treatment with fluoxetine or quinidine (two potent CYP2D6 inhibitors) or placebo
No significant difference among the three groups in daily codeine intake or depression scores. By end of treatment, large decrease from baseline in mean daily codeine use in all groups: placebo by 57%, quinidine by 56%, and fluoxetine by 51%

Low

Frei 201019

Case series, Australia

27 (48% female); 20 years or more

Opioid pharmacotherapy (16 patients), buprenorphine taper (3), buprenorphine maintenance (10), methadone (3). Outcomes not reported

Very low

Nilsen 201067

CBT-based clinical trial with partial or complete codeine reduction, Norway

11 (82% women); mean, 43 years

Two specifically trained physicians delivered six CBT sessions (verbal reattribution, behavioural experiments), tapering codeine gradually within 8 weeks
Codeine use significantly reduced from mean 237 mg to 45 mg; six ceased codeine

Very low

Thekiso 201042

Retrospective chart review, Ireland

20 (65% female): mean, 49.2 years (SD, 23.4 years)

Treated for substance withdrawal with standard pharmacological protocol-driven regimes, underwent up to 4 weeks’ comprehensive inpatient treatment. Withdrawal regime included tapering benzodiazepines and clonidine. Affective comorbidities also treated (pharmacological, “psycho-education”
Mean length of stay, 42 days; mean length of treatment for withdrawal, 16 days

Very low

Cooper 201338

Qualitative interviews, United Kingdom

25 (52% female); range, 20–60 years

One-quarter received drug treatment (methadone, buprenorphine) from treatment service or GP. Online support important for attempts to self-treat
buprenorphine and methadone often achieved to positive outcomes — either on maintenance doses or opiate free — despite initial reservations

Very low

Nielsen 201527

Retrospective chart review, Australia

135 (53 codeine-dependent, 66% women; mean, 38.6 years)

Codeine dependence more likely to be treated with buprenorphine than methadone, withdrawal management more common than longer term pharmacotherapy. Outcomes not reported

Very low

Nielsen 201543

Retrospective chart review, Australia

19 (84% female); mean, 41 years (SD, 9 years)

Buprenorphine maintenance treatment by drug treatment services, five as inpatients,14 as outpatients
Median dose, 12–16 mg buprenorphine, four patients continued to use opioids. buprenorphine doses higher than estimated based on codeine dose

Very low

Van Hout 201534

Qualitative interviews, Ireland

21 (57% female): mean, 39 years (range 26–62 years)

Methadone (14 patients), buprenorphine (3). Supportive medical care and a slow tapering of codeine products or substitution.
Buprenorphine viewed particularly positively in removing craving and withdrawal effects. Relapse with codeine tapering was common; attributed to lack of effect on cravings and use of over-the-counter codeine

Very low

Norman 201669

Mixed methods (systematic review, qualitative interviews), Ireland, United Kingdom, South Africa

23 interviews with key experts

Buprenorphine and methadone in substitution therapy. Notes efficacy of CBT for treating opioid dependence
Outlined “best practices” in treatment reported by stakeholders, suggested “innovations” for treatment. Did not assess treatments

Very low

CBT = cognitive behavioural therapy; CYP = cytochrome P450; SD = standard deviation. For further details, see online Appendix, table 4D.

Authors

  • Suzanne Nielsen1
  • Tim MacDonald2,3
  • Jacinta L Johnson4,5
  • 1. National Drug and Alcohol Research Centre, University of New South Wales, Sydney, NSW
  • 2. Currumbin Clinic, Gold Coast, QLD
  • 3. Griffith University, Gold Coast, QLD
  • 4. University of South Australia, Adelaide, SA
  • 5. Southern Adelaide Local Health Network, Adelaide, SA

Correspondence: suzanne.nielsen@unsw.edu.au

Competing interests

Suzanne Nielsen is a named investigator on untied educational grants from Reckitt–Benckiser and Indivior. Tim MacDonald has received honoraria, fees and professional development resources from Servier, the Australian and New Zealand Mental Health Association, and Healthe Care; he works in the private sector and receives income for clinical services.

Acknowledgements

Suzanne Nielsen holds a National Health and Medical Research Council Research Fellowship (1132433).

References

Linked content

  • MJA InSight: Hidden codeine dependency set to reveal itself

  • MJA Podcast: Dr Suzanne Nielsen

Provenance: Not commissioned; externally peer reviewed.

View this article on Wiley Online Library

More like this

View more