Use of out-patient commitment

Incidence and point prevalence rates found in the present study are higher than rates reported from England,^{Reference Trevithick, Carlile, Nodiyal and Keown14} but lower than most jurisdictions in New Zealand and Australia, where rates are reported to be above 100 per 100 000 population at risk.^{Reference O'Donoghue, Brophy, Owens, Rasic, McCullough and Huang3,Reference O'Brien15,Reference Light, Kerridge, Ryan and Robertson16} This cannot be explained by demographic differences, as our study sample shows the same characteristics as found in most other out-patient commitment studies. However, reservations should be made when incidence and prevalence rates between jurisdictions are compared, because reference populations vary in reports. Some rates are based on total populations, others on adults only and some are based on adults up to the age of 65.^{Reference O'Donoghue, Brophy, Owens, Rasic, McCullough and Huang3,Reference Trevithick, Carlile, Nodiyal and Keown14}

Patients were middle aged (30–50 years) with a slight male predominance (56.4%).^{Reference Burns, Rugkåsa, Molodynski, Dawson, Yeeles and Vazquez-Montes4,Reference Rugkåsa, Nyttingnes, Simonsen, Benth, Lau and Riley6,Reference Churchill, Owen, Singh and Hotopf17} The percentage of all involuntary in-patients who received an out-patient commitment order at discharge varied between 9.2 and 11.9% over the study years, which is about the same level as reported from England,^{Reference Trevithick, Carlile, Nodiyal and Keown14} but lower than other jurisdictions.^{Reference Burgess, Bindman, Leese, Henderson and Szmukler18} The percentages of involuntary in-patients discharged on out-patient commitment orders as well as incidence and prevalence rates differed significantly between sites in the present study. The finding is not surprising as geographical variations within jurisdictions for both in-patient commitment and out-patient commitment are more the rule than the exception.^{Reference O'Brien15,Reference Light, Kerridge, Ryan and Robertson16,Reference Weich, McBride, Twigg, Duncan, Keown and Crepaz-Keay19} The mechanisms underlying such differences are not fully understood, and cannot be explained by demographics, disorder severity or service structures alone.

A general problem related to reports on out-patient commitment incidence and prevalence rates is the varying quality of administrative data. In Norway and Denmark, discrepancies have been found between data recorded by the health authorities and research data based on patients' files. Research data usually identified more cases than those reported by health authorities.^{Reference Iversen, Høyer and Sexton20,Reference Poulsen21} A strength of the present study is that data on incidence and prevalence rates are complete, because we were allowed to access all patients' medical files without the consent of individual patients, and data on out-patient commitment in the files were rarely missing. In addition, all data were thoroughly verified by the research team. Thus, we feel confident that the data reported in this study are well suited as a reference point for developments in the use of out-patient commitment, as well as comparisons with other jurisdictions on the use of out-patient commitment.

It is impossible to draw any firm conclusions as to whether a rational use of out-patient commitment is reflected in the proportion of in-patients discharged on an out-patient commitment order, or the incidence and prevalence rates found in this or other studies. To evaluate the use of out-patient commitment orders, one needs to consider the purpose of such orders. Out-patient commitment has different and often ambiguous justifications in different countries. Justifications for the introduction of out-patient commitment include efforts to reduce in-patient care, to increase after-care adherence, to reduce relapse, to prevent revolving-door regimes, to minimise coercive care, to prevent violence or a mixture of these aims. In this context, factors such as the coercive power of out-patient commitment orders, the expressed aims of such orders and how out-patient commitment works in practice, need to be considered. Likewise, the quality and availability of out-patient mental healthcare and social services are important factors when assessing the rationality and efficacy of out-patient commitment regimes.^{Reference Meldrum, Kelly, Calderon, Brekke and Braslow1}

Duration of out-patient commitment orders and time to readmission

The median duration of out-patient commitment in our study (354 days) is close to what was found in a study from England that followed up patients on community-treatment orders over 36 months (364 days).^{Reference Burns, Yeeles, Koshiaris, Vazquez-Montes, Molodynski and Puntis22} In both studies, the mean duration was longer, reflecting a relatively small group of patients on long periods of out-patient commitment. The determinants of longer out-patient commitment duration found in the current study indicate that patients with a longer history of severe mental illness are likely to spend a longer time under out-patient commitment orders. The differences between sites in duration of out-patient commitment orders may reflect local differences in treatment cultures and service provision. Qualitative interviews with clinicians responsible for patients on an out-patient commitment in Norway revealed different responses to patients who seemed to benefit from the out-patient commitment regime: some clinicians underlined the importance of continuing out-patient commitment for patients who did well, whereas others said they discontinued out-patient commitment because the patients had improved and could be managed without mandated treatment.^{Reference Riley, Lorem and Høyer23} The impact of treatment culture and ideology on the use of out-patient commitment is little researched, and future studies on this area should be encouraged.

Our findings that many in-patient episodes before out-patient commitment predicted shorter time to readmission, and that using antipsychotic medication during the whole out-patient commitment period predicted longer time to readmission, could be explained by the same phenomenon, namely adherence to medication, but we are unable to draw any reliable conclusion based on the study data. Interaction between in-patient care, out-patient commitment and voluntary out-patient care is complex, and influential factors might have been omitted from our study. Differences between sites in time to readmission are also likely to reflect local differences in treatment cultures and service provision.

Outcome of out-patient commitment orders

The most common outcome measure used in assessing out-patient commitment regimes is changes in the use of in-patient services before and after out-patient commitment treatment.^{Reference Barnett, Matthews, Lloyd-Evans, Mackay, Pilling and Johnson2,Reference Kisely and Hall9} In our study we found no reduction in number of admissions, but significant reductions in total number of in-patient days (P<0.001) and number of days per admission (P = 0.043) when comparing 3 years before with 3 years after implementation of a first-ever out-patient commitment order. The number of readmissions was also reduced, but not significantly (P = 0.092). However, all randomised controlled trials on out-patient commitment programmes have found no evidence that out-patient commitment regimes reduce the use of in-patient care.^{Reference Burns, Rugkåsa, Molodynski, Dawson, Yeeles and Vazquez-Montes4,Reference Kisely and Hall9–Reference Swartz, Swanson, Wagner, Burns, Hiday and Borum11} In this respect, reservations should be made since the before/after design applied in this study is less robust than randomised controlled designs. Nevertheless, we believe that use of in-patient services as an outcome measure can be questioned regardless of study design. It makes a considerable difference whether the readmission is at the request of the patient or involuntary. Many patients subjected to out-patient commitment have a long psychiatric history and many have poor social supports and networks.^{Reference Rugkåsa, Nyttingnes, Simonsen, Benth, Lau and Riley6,Reference Churchill, Owen, Singh and Hotopf17} Patients may sometimes feel a need for a more protected and safer environment that can be offered by regular out-patient services. Crucial factors in this context are that patients are in control of when they can access and leave in-patient services and that in-patient treatment is on a voluntary basis, despite being under an out-patient commitment order. Such user-initiated in-patient periods may be perceived in a positive light, unlike coercive readmissions to in-patient care. A small-scale study from Norway^{Reference Heskestad and Tytlandsvik24} on the effect of letting patients with long-term psychotic disorders and comorbid substance misuse decide on in-patient care themselves, is relevant in this context. Despite shortcomings in study design, such as the selection of patients authorised to make self-determined admissions, the outcome was noteworthy: the number of readmissions went up, whereas the length of stay, both per admission and in total, went dramatically down (mean 2.5 days). Most important, involuntary admissions were reduced by 50%.

The question of readmission as an outcome measure in out-patient commitment studies was raised in 2001,^{Reference Høyer and Ferris25,Reference Draine26} but the use of in-patient care is still the most common outcome measure in quantitative studies on out-patient commitment regimes. In the present study, we attempted to record voluntary versus involuntary readmissions, but such data were either lacking or unclear in many cases, which meant that we had to exclude the voluntary/involuntary distinction from the analysis. We are not aware of any outcome studies of out-patient commitment that have distinguished between voluntary and involuntary readmissions. Research incorporating this distinction can provide a better understanding of the impact on in-patient care from a voluntary–involuntary perspective.

Pattern of out-patient commitment use

All patients in our study with a first-ever out-patient commitment order (n = 274) were transferred to out-patient commitment regimes directly after discharge from an in-patient stay. Unfortunately, we did not record whether this applied to the whole sample (n = 1414). Our data collectors, however, could not remember any cases of out-patient commitment orders that did not directly follow discharge from an in-patient period. This finding was surprising given the heated public debate when the possibility of making an out-patient commitment order without any prior in-patient period was introduced in 1999. Opponents of the proposal believed that this option represented a widening of the coercive net. Apart from a review from the USA reporting that eight states have the option to place patients on out-patient commitment without any prior in-patient period,^{Reference Meldrum, Kelly, Calderon, Brekke and Braslow1} we are not aware of other jurisdictions where direct out-patient commitment placement is a legal option. Why clinicians may refrain from using this possibility needs to be explored in future research.

Out-patient commitment and involuntary treatment

Just under half of the first-time out-patient commitment patients received an involuntary treatment order, a requirement under the Norwegian Mental Health Act for the treatment of involuntarily placed patients without valid consent. Further, the frequency of involuntary treatment orders varied significantly between sites, from 19.3 to 77.5%. As patients report in qualitative studies, the most controversial issue concerns medication regimes, where patients challenge the kind of medication, dosage, duration and method of drug administration.^{Reference Riley, Høyer and Lorem12,Reference Corring, O'Reilly and Sommerdyck13} Why so many patients on an out-patient commitment seem to comply with the medication prescribed without being subject to a separate involuntary treatment order is an interesting question. One possible explanation might be that patients on an out-patient commitment usually have long-term experience of mental health services, including the use of involuntary measures. This has taught them that if they resist taking the prescribed medication, an involuntary treatment order would be made, which could place them in a worse situation than if they comply.^{Reference Riley, Høyer and Lorem12} If this holds true, it can be questioned whether patients on an out-patient commitment without an involuntary treatment order take their medication voluntarily based on free consent. Norway is one of the few countries in Europe where a separate order for involuntary treatment is required in addition to the involuntary placement order. The obligation to make separate decisions for involuntary placement and treatment is intended to be an additional legal safeguard for interventions that may represent a serious violation of patients' integrity. The fact that so many patients on an out-patient commitment in this study do not have an involuntary treatment order, and thus miss the opportunity to legally challenge involuntary treatment, is a matter of concern from a patients' rights and patient autonomy perspective.

Implications

Determining the factors resulting in the pronounced variations in the use of out-patient commitment between sites calls for further study. When considering readmissions as an outcome measure in out-patient commitment studies, it may be important to distinguish between voluntary and involuntary in-patient placement. Also, the reason why only fewer than half of the patients with an out-patient commitment had a separate involuntary treatment order needs to be studied in more detail.