Gender differences in reoffending after discharge from medium-secure units: National cohort study in England and Wales

Background Previous research has shown that there are gender differences in reoffending after discharge from medium-secureunits, but these have not been adequately explained.

Aims To investigate gender differences in reoffending after discharge from medium-secure psychiatric units.

Method All people discharged from medium-secure units in Englandand Wales between April 1997 and March 1998 were followed upfor 1 year (n=959; 12% women). Reoffending was estimated bycollecting reconviction data from the Home Office's Offenders'Index or from files at the mental health unit up to 2 yearsafter discharge.

Results Women were less likely than men to be reconvicted within2 years of discharge (9% v. 16%, OR=0.49, 95% CI 0.25-0.98). Adjustments for history of self-harm, drug or alcohol problemsand previous offending substantially reduced the gender difference.In the full model the OR was 0.97 (95% CI 0.45-2.12).

Conclusions Some or all of the gender differences in reoffending between men and women are explained by self-harm, alcohol anddrug problems and previous criminal history.

INTRODUCTION

TOP

INTRODUCTION

Previous studies of psychiatric patients discharged from medium-secure psychiatric units have suggested that women are less likelyto reoffend after discharge compared with men (Coid et al, 2000;Edwards et al, 2002). However, results have been inconclusivebecause of the comparatively small numbers of women. Lowerrates may be owing to the confounding effect of variables frompsychiatric or forensic history that are known to differ betweenmen and women, such as a history of self-harm (Hawton et al, 2003), physical and sexual abuse (Edwards et al, 2003), drug or alcoholproblems (Farrell et al, 2001) and previous criminal history(Maden et al, 2004). It is important to know whether these variables explain the gender differences in reoffending becausethis may help to identify patients with lower or higher riskfor reoffending.

In a previous paper (Maden et al, 2004) we examined the incidenceand risk factors for reoffending in the whole sample. Thispaper examines data from the same national cohort study tosee whether differences in reoffending between men and womenexist and if any differences persist after having taken intoaccount the possible confounding factors.

METHOD

TOP

METHOD

Description of the data-set
We used data from the Pathways out of Medium Security studycommissioned by the Department of Health (Maden et al, 2004).This study included all individuals (n=959) discharged frommedium-secure units in England and Wales between 1 April 1997and 31 March 1998: 843 men (88%) and 116 women (12%). The samplewas drawn from 34 units, 28 within the National Health Service(NHS) and 6 independent units. Further details about the aimsand method of the original study are available on request fromthe authors.

Ethical approval for the main study was obtained from SouthThames Medical Research Ethics Committee.

Measures
Socio-demographic and clinical variables
Socio-demographic variables and clinical characteristics ofthe sample were obtained from the patient notes. We collectedinformation on the following variables: age at admission, sourceof referral, main diagnosis, history of previous admissionsto psychiatric hospital, history of physical or sexual abuseduring childhood/adolescence, history of self-harm, historyof drug or alcohol problems.

Follow-up data
Follow-up location data, including readmission to a psychiatrichospital, were collected for a period of 12 months after dischargeor transfer by writing to the consultant who took over carewhen the person was discharged or transferred. When patientshad been transferred to other hospitals, data were collectedfrom the receiving hospital's medical records department.

Forensic data
Background data in relation to the index offence were collectedfrom the medical records department at each unit. We also recordedthe legal status of the admission (voluntary or involuntary)and the number of previous convictions. Reoffending was assessedby collecting information on reconvictions from the Offenders'Index at the Home Office. Data collection time was extendedto 2 years for reconviction data because some offences may takemany months to go to a court, and then there is a further delaybetween conviction in a court and recording of this informationin the index. However, less-serious offences committed by peoplewho were not sent to court were not recorded. In addition,all convictions that appeared in the Index the first 6 monthsafter discharge were manually checked to ensure that they were referring to new offences committed after discharge and notto the index offence.

Data analysis
All data analyses were conducted using Stata version 7.0 forWindows. A non-parametric non-parametric kappa sample testfor the equality of medians evaluated differences in the ageat admission and length of stay between women and men. Pearson's ${chi}$ ² test was used to test for gender differences in the variousadmission, clinical and forensic variables. The associationbetween gender and reconviction was estimated with a seriesof logistic regression models using the logit command in Stata.We used the robust option of the logit command combined withthe cluster option, to take into account the clustering ofthe observations within the medium-secure units. We used reconvictionat either the first or the second year (Yes/No) as the dependentvariable. Crude odds ratios (with 95% confidence intervals)were first obtained for gender; then we entered into the modelage and history of self-harm, physical and sexual abuse, alcohol-and drug-related problems and the number of previous convictions.A final model adjusted for all these variables.

RESULTS

TOP

RESULTS

Demographic and admission characteristics
Table 1 shows gender differences in admission characteristicsand other clinical variables. Women had higher historical levelsof self-reported physical and sexual abuse and much higherlevels of self-harm than men. Women, however, were less likely than men to have a drug problem and possibly less likely tohave problems with alcohol, although the latter did not reachstatistical significance. Women were more likely than men tobe admitted with a personality disorder and were more likelyto have had previous psychiatric treatment. No difference was found between men and women in terms of re-admission to anypsychiatric hospital over follow-up.

View this table:
[in this window]
[in a new window]

Table 1 Gender differences in admission characteristics, clinical variables and clinical course of 959 patients discharged from medium-secure units

Forensic characteristics
Table 2 shows details of gender differences in various forensicvariables. The forensic profile of men differed significantlyfrom that of women. Men were more likely to be referred fromprison, and the index offence was more likely to concern propertyor be of a sexual nature. Men were also more likely to havetwo or more previous convictions and/or previous prison sentences.

View this table:
[in this window]
[in a new window]

Table 2 Gender differences in forensic-related variables of 959 patients discharged from medium-secure units

Association between reconviction and gender
Table 3 shows that women were less likely to be reconvictedcompared with men and the crude OR was 0.49 (95% CI 0.25-0.98),that is women were half as likely to be reconvicted comparedwith men. We hypothesised that a number of variables would reducethe gender differences in reconviction, and these models arealso presented in Table 3. Adjustment for self-harm and numberof previous convictions had the strongest effect. In the finalmodel, adjustment for all variables reduced significantly thegender differences in reconviction (OR 0.97, 95% CI 0.45-2.12).In this model significant independent predictors of reconvictionwere age, self-harm, history of drug problems and number ofprevious convictions.

View this table:
[in this window]
[in a new window]

Table 3 Odds ratios for reconviction in 116 women compared with 843 men discharged from medium secure units

DISCUSSION

TOP

DISCUSSION

Summary of main findings
This study found that women were less likely to be reconvictedthan men, but adjustment for a number of variables substantiallyreduced the gender difference. Logistic regression analysisshowed that the gender difference in reconviction was partlyexplained by the increased level of self-harm in women, togetherwith less previous offending and reduced drug problems in comparisonwith men. This study also found significant univariate differences between men and women in a number of demographic, clinicaland forensic variables, with women being less criminal andmore likely to have a past psychiatric history than men.

Comparison with other studies
One of the key findings from the literature review on womenand secure psychiatric services (Lart et al, 1999) was that,although women make up less than one-fifth of the populationin secure settings in Britain, they are a heterogeneous group,with a wide range of ages and personal, psychiatric and forensichistories. This was found to be the case in this study. Itwas also observed in the 1999 review that women in secure psychiatricservices have a different pattern of diagnosis compared withmen; in particular, in the medium-secure services they aremore likely to be diagnosed as having a borderline personalitydisorder than their male counterparts. Again, this was confirmedin our study.

Lart et al's (1999) review notes the absence of recent outcomestudies including women. In high-secure hospital samples, Buchanan(1998) found that gender exerted no independent effect, andJamieson & Taylor (2004) also showed that there was nostatistically significant difference between men and women inthe proportion of each reconvicted.

Self-harm and physical or sexual abuse
In terms of levels of self-harm and abuse, there are few studieswith which to compare our data as there is even less clinicaldetail available on women in medium security than on womenin high security. Bland et al (1999) described 87 women in Broadmoor in 1994 and found that nearly 70% had a confirmedor suspected history of childhood sexual abuse and 94% hada history of self-harm. Heads et al (1997) found that forwomen with schizophrenia in special hospitals, rates of childhood sexual and physical abuse were significantly higher than fortheir male counterparts. A history of self-harm and sexualabuse is more likely in women (Lart et al, 1999). Historiesof early physical or sexual abuse are particularly common inadults with a diagnosis of borderline personality disorder,and may represent a final common pathway for future impulsiveand aggressive offending behaviour (Ogata et al, 1990; Shearer et al, 1990). A compulsion to repeat early trauma may be a manifestationof the re-experiencing phenomena of post-traumatic stress disorder (Deblinger et al, 1989). Physical abuse and sexual abuse mayincrease the risk of violence against others, whereas self-harmis violence against one's self and may lower the risk for violenceagainst others. Self-harm has been shown to be associated witha lower risk of reconviction (Maden et al, 2004), whereasa history of sexual abuse has been shown to be associated witha higher risk (Maden et al, 2004).

Alcohol and drug problems
Further, alcohol and drug problems are more common in men andsubstance misuse has been shown to be associated with an increasedrisk of reconviction (Maden et al, 2004; Scott et al, 2004). Bland et al (1999) found that 38% of their high-secure hospitalsample had an alcohol problem and 37% had a drug problem, levelsvery similar to those found in our study. There is a well-establishedlink between substance misuse and higher rates of violenceby people with major mental illness (Arseneault et al, 2000).In a medium-secure hospital sample, Baxter et al (1999) foundthat comorbidity, with conduct disorder or problem alcoholuse, doubled reoffending compared with schizophrenia alone,whereas young age or polydrug use or conduct disorder increasedreconviction rates by factors between 2 and 3.

Previous convictions
It has been shown that the strongest predictor of reoffendingis the number of previous convictions (Bowden, 1981; Black, 1982) and the current study confirmed that. In our own study womenappear to have a lower risk of being reconvicted because theytend to less often have a history of previous convictions orof drug problems, and more often have a history of self-harm.

Limitations of the study
The findings of the present study should be considered in thecontext of the following limitations. First, we were not ableto record all types of new offences but only those that ledto conviction. Therefore our results cannot be applied to peoplecommitting minor offences. This may underestimate the realimpact of antisocial behaviour in both men and women. Second,data on reconviction were only obtained from the Offender'sIndex at the Home Office and this will be inaccurate. The intervalbetween committing an offence and being convicted of it ina court is often many months (especially in the case of thosewith mental illness), and there is further delay between conviction in a court and recording of this information in the Index.In order to minimise this misclassification, we extended theperiod of data collection for 2 years regarding reconviction.In addition, all convictions that appeared in the Index withinthe first 6 months of discharge were manually checked, to ensurethat they were referring to new offences committed after dischargeand not to the index offence. For practical reasons, we werenot able to collect information from other sources such asthe national police computer records. In any case, we thinkthat any misclassification would be more likely to bias theresults towards the null value, i.e. to further reduce the gender difference in reoffending. Third, medical records, which wereusually kept on the units themselves, were the main sourceof information related to the index admission. They provideddetails of demography, medical and offending history, sourceof referral, reason for referral, diagnosis and destinationon discharge. Since the information was not collected for thepurposes of this particular study, a degree of measurementbias is inevitable, particularly regarding information on self-harm,physical and sexual abuse and drug and alcohol use. If thisbias was not random it could influence the results in eitherdirection. Last, even in this large cohort of 959 patients,the number of women was relatively small and the power of thestudy may have been compromised.

The government is taking a wide-ranging approach to tacklingthe inequalities that affect women. The Department of Health's (2002) publication Women's Mental Health: into the Mainstreampoints out in regard to secure and forensic services that thereare differences in the social and offending profiles of womenand men, their experience of mental ill health, their patternsof behaviour, and their care and treatment needs. Our study highlights these differences and shows that it is possibleto use a number of clinical and forensic variables to assessthe risk of future offences in individuals discharged frommedium-secure units in the UK. Future studies should try toaddress whether interventions aimed at reducing the impact of self-harm or abuse in women and of alcohol or drug problemsin men could lower the risk of reoffending.

REFERENCES

TOP