We had hoped that our article would stimulate a balanced discussion about this complex issue. We entirely agree with the view expressed in both letters that trust and therapeutic alliance between the patient and the treating team are critical components of the recovery process. We do not believe, however, that use of electronic monitoring necessarily indicates a lack of trust. It was envisioned that the device be used primarily for patients in the initial stages of taking leave as part of their clinical pathway towards discharge into the community. Our clinical experience, supported by as yet unpublished data, confirms that this has been the case in our service. In these circumstances, electronic monitoring may even help to further develop a trusting relationship between the wearer and the team, by granting earlier and more frequent leave and by allowing the patient to demonstrate avoidance of exclusion zones when on unescorted leave. There must be a balance between trust and therapeutic optimism in our treatment of our patients. Furthermore, viewing trust as being simply ‘present’ or ‘absent’ would be a naive approach in forensic services. These questions are being explored in quantitative and qualitative research of electronic monitoring in our service.

Both letters raise concerns about granting of leave for high-risk patients. Watson et al point out that decisions surrounding leave are complex, a view that we share. However, the implied view in both letters that patients can be discretely classified into high risk for absconding or not is again overly simplistic. Clinical impression alone in risk assessment has been shown to be unreliable and validated risk assessment tools have been shown to be more useful in identifying individuals at low rather than high risk. ^{Reference Fazel, Singh and Doll1} No validated tool for the assessment of absconding risk yet exists, though we are currently working on developing one. Risk management, therefore, involves a component of positive risk-taking aided by creative management strategies. We propose that electronic monitoring is such a strategy.

Watson et al are liberal in their use of the term ‘coercion’. A policy was put in place whereby patients were informed that use of electronic monitoring was optional and if they chose to decline to wear the device, their leave would be risk assessed as per normal procedure. Consent is another complex issue in psychiatry and can be defined in degrees, rather than as a binary concept. ^{Reference Konow2} It is true that patients’ decisions about consent to electronic monitoring are likely to be influenced by their wish to move more quickly towards leave and discharge. This has parallels with consent to medication and engagement in psychotherapies and occupational activities, particularly in the forensic setting.

Watson et al express concern about forensic services being closely aligned with the prison system. We believe that the use of secure units with locked wards and secure perimeters represents a level of coercion much more closely aligned to this system than does electronic monitoring. Any strategy that can help minimise the amount of time spent in such units would then surely be a welcome development for those concerned about patient liberty and overall progress. Far from making our units more like prisons, one of the key aims of our strategy was to allow for engagement in community leave and activities at the earliest possible stage. As Simpson & Penney point out, electronic monitoring may allow the person more apparent personal freedom than their clinical risk would otherwise allow.

The article referenced in The Sun was chosen as an example of media coverage of such absconding events. That such reports are often sensationalised or biased is one of the many challenges facing mental health services and patients. Media coverage of absconding events leads to reputational damage for services and can undermine the confidence of the community. We cannot and should not ignore community attitudes towards system breaches, especially as clinicians will be held to account when they occur. Another of our aims is therefore to reduce the frequency of these incidents, for the protection of the public and the reputation of our service.

Watson et al are correct in saying that electronic monitoring cannot directly prevent violent incidents. We believed that this was self-evident and therefore we did not address this issue in our article. Regarding costs, a cost-benefit analysis is currently underway. As our article states, our service was acutely aware of the important ethical considerations and we sought legal and ethical advice. A commentary addressing legal and ethical issues in more depth is currently being prepared. The questions Simpson & Penney raise about reoffending, recovery and longer-term outcomes are valid and we hope to address these in our future research.