Observation of People with Mental Health Problems

A revision of the 1995 CRAG document Nursing Observation of Acutely Ill Psychiatric Patients in Hospital to bring it in line with current clinical practice and policy terminology.


4 RISK (Assessment, decision making and recording)

4.1 The decision to use an increased level of observation is based on a variety of factors. Central to it must be the risk assessment of that patient's mental state at that moment in time. Risk assessment/management is a complex process involving both objective data (such as patient history, behaviour, etc.), data from third parties, and the judgement of the clinicians involved. The recently published report "Risk Management" (2000) offers comprehensive guidance on this subject and also acknowledges that the process is difficult and may have to be repeated frequently if the patient's clinical state is fluctuating.

4.2 It is important to note that in the original report it was observed that:

"Few, if any, properly validated risk assessment tools are available and the group sees an urgent need to conduct audit and research into the most clinically effective ways of assessing risk and prescribing the appropriate level of nurse observation."

It is relevant to note that this position, in many ways, has not fundamentally changed in that any value of formal rating scales and check lists must be tempered by the central role of clinical judgement of an experienced professional. Simple checklists of questions/prompts are valid in guiding less-experienced staff in carrying out a risk assessment. An example is given in the "Risk Management Report" and is included in Annex 5 of this report. The key issue in sound risk assessment is that there is open and in-depth dialogue between all members of the clinical team and with appropriate others including relatives, carers and the patient themselves and that risk assessment is a dynamic process that requires constant review.

4.3 Ideally a multi-disciplinary team should always make these decisions. However, on many occasions (particularly at weekends and evenings) decisions may have to be made by a doctor and the ward nursing team. Such decisions should always be reviewed at the first available opportunity with a larger number of the full team.

4.4 Local policies must clarify the procedure for increasing levels of observation in emergencies. This decision should be able to be made by the senior nurse in charge of the unit on their own initiative but followed up by consultation with appropriate medical staff as soon as possible. Staff must feel empowered to raise levels of observation and be supported in this action (even if this increase is subsequently reduced following a broader team discussion). Teamwork and trust between team members are essential to safe decision making and safe practice.

4.5 The reduction in the level of observation should ideally be a team decision. To ensure patients are not left on an increased level inappropriately it is recommended that teams plan ahead, particularly at weekends, clarifying the circumstances that would enable a reduction in observation level. As part of this process there should be a clear local policy on the acceptability of the authority of the nurse in charge to reduce observation levels. There should be a generic policy at Trust level, which sets out the broad principles including a clear statement of support for nursing staff implementing these decisions. There must also be a specific plan for each patient, which outlines the agreed changes in behaviour that would facilitate a reduction in observation level and the exact procedure for this decision to be actioned. It must detail the role of duty medical staff or senior nurses in this process. It may be appropriate for the policy to differentiate between the procedure for the reduction of the observation level from special to constant compared to a reduction from constant to general.

The following guidance from the Royal College of Psychiatrists clarifies the role of medical staff in the decision-making process regarding levels of observation:

Deciding on observation levels should be a process in which the multi-disciplinary teams should all contribute. The Scottish Division of the Royal College of Psychiatrists state (consultation response, January 2002) that they see no problem in non-medical members of the team being able to reduce observation levels as long as (the circumstances which would allow) a future reduction in observation levels could be specified at (multi-disciplinary) ward rounds and documented in the case notes in advance. Nursing staff usually know the patients better and are in quite a strong position, for instance, at weekends to decide if observation levels can be reduced. The on-call junior doctor should be involved in examining the patient's mental state and discussing the risk with the senior nurse on duty before the decision to reduce observation is made. Where there is any doubt the decision should be discussed with the on-call consultant or postponed until the next opportunity when the full team, including the patient's Responsible Medical Officer (RMO), is present. The RMO retains final responsibility.

4.6 There should always be a record of decisions regarding observation kept within the patient's notes including an explanation as to why an increased level is used. It is recommended that a simple record is kept to allow auditing of the frequency, level and duration of increased levels of observation as well as the clinical reason(s) behind the choice. The record must clearly show the perceived risks which led to the decision, who was involved in the decision, and the patient's opinion of the need for increased observation. This audit trail provides key information both in monitoring the frequency of the usage of raised levels of observation and in Critical Incident Reviews (see paragraph 12.3).


5.1 One of the key areas of clinical practice in acute psychiatric care is deciding what intensity of care is needed for individuals. The original document refers to three levels:

  • general

  • close

  • special.

In three NHSScotland Primary Care Trust policies we examined, reference is made to the use of a fourth level. This involves the use of a specified time period (such as every 15 minutes) at which the whereabouts of the patient must be "checked". The group gave this issue considerable debate and sought views from outwith the group as it is clearly a contentious issue and one on which a clear unambiguous statement must be given.

5.2 A summary of the points for and against is given below.


  • allows an intermediary level between intense one-to-one observation and general observation (particularly when reducing the level of observation)

  • is less intrusive for person being observed

  • is less staff-intensive

  • may be used to comply with general observations requirements.


  • high risk as patient able to carry out risk behaviour during gaps in observation

  • maybe used as "easier" option when constant observation is really indicated

  • encourages a mechanistic process of care

  • does not fulfil the purposes of observation.

5.3 The group decided that the original three levels of observation are still the most appropriate and that on balance the risks inherent within "timed checks" outweigh the possible benefits. This view is upheld within "Safety First" (5-Year Report of the National Confidential Inquiry into Suicide and Homicide by People with Mental Illness, 2001) which highlights the risks inherent within "timed" or what they term "intermediate" observation. It recommends that "alternatives to intermediate level observations be developed for patients at risk" referring to intermediate-level observations as "of unproven benefit". Being aware of the patients' whereabouts supports good nursing practice but should not be considered part of the observation process.

5.4 The "timed check" form of observation is seen by the group as unsafe and should not be used as a means of meeting a need for an increased level of observation.

5.5 It was clear that the majority of Trusts and staff consulted considered that the original recommendations regarding levels of observation were well implemented, useful and well integrated into the care package; comments received by the group during consultation were supportive of this view. There seemed no requirement therefore to alter terminology or the major aspects of clinical practice. The three levels of observation suggested in the original report remain valid and should continue to form the basis of local policy.

5.6 With this in mind, we continue to recommend that the following categories of observation be used:




5.7 General Observation

The general level of observation is intended to meet the needs of most patients for most of the time. It should be compatible with giving patients a sense of responsibility for their use of free time in a carefully planned and monitored way. The staff on duty should have knowledge of the patients' general whereabouts at all times, whether in or out of the ward. This could be achieved by establishing a patient allocation system whereby the nurse in charge is kept informed of each patient's whereabouts. Patients on general observation are considered not to pose any serious risk of harm to self or others and are unlikely to leave the ward area or other treatment departments without prior permission, escort, or at least informing staff of their planned destination. Any limits set should be determined in conjunction with the patient, documented and updated in the care plan as necessary.

5.8 Gournay & Bowers (2000) offer a useful description of general observation:

"General Observation can be thought of as the observation and monitoring of the physical geography of the ward and as a component of constant review of safety in the light of the opportunities the ward and its contents provide for harm to come to patients. This general observation should be an established part of the ward routine and followed rigorously and regularly by nurses, as part of their everyday practice to maintain the safety of the patients."

5.9 Constant Observation

The constant level of observation should be used for patients considered to pose a significant risk to self or others. An allocated member of staff should be constantly aware at all times of the precise whereabouts of the patient through visual observation or hearing. The method and purpose of maintaining observation must be clearly determined and stated at the time of review. Respect for privacy should be an important consideration, but a balance should be struck on the side of safety in all matters such as escorting to the toilet, bathroom, or public telephone, etc. In some circumstances the patient may be permitted to leave the ward or other clinical area in the company of an escorting nurse, other informed professional worker or appropriate relative. This decision must be part of the risk assessment process and the comments referred to in the previous section should be noted. Appropriate members of the multi-disciplinary team (generally a minimum of the nurse in charge and duty doctor) should review the need for constant observation at least every 24 hours.

5.10 Special Observation

The special level of observation will generally be rarely prescribed. The patient should be clinically assessed as requiring intensive and skilled intervention as a consequence of their very serious mental and/or physical state. The patient should be in sight and within arm's reach of a member of staff at all times and in all circumstances. Considerations of privacy would be subordinated to those of safety. In some situations more than one staff member may be required. In the event of the patient leaving the ward an appropriate number of escorts should accompany the patient. As this form of observation is potentially very intrusive, it should only be used when judged strictly necessary by the clinical team, and this level of observation should be subject to frequent review (at least every 24 hours) involving appropriate members of the team. A system should be in place for dealing with the increased demand on staff resources which special observation creates. Only staff familiar with the condition of the patient on special observation should normally be deployed on this demanding work.


6.1 It must be acknowledged that it is primarily psychiatric nurses who provide 24-hour care and who will, therefore, carry the majority of the responsibility for the observation of the patients. However, with the emphasis on multi-disciplinary team-working and the increasing role of users and carers, consideration should be given to the role of these groups within this area of care. It seems correct that in appropriate situations other professionals (apart from nurses) should be involved and have responsibility for the observation of a patient. Indeed, it is clear from our consultation that this practice is already in use to some degree and that guidance on it would be welcomed.

6.2 For non-nursing staff to be involved in observation the following issues must be addressed:

  • there must a fool-proof system of staff knowing who is responsible for the observation of a patient at all times

  • there must be a simple way of communicating between staff members all changes in the level of observation

  • all staff must accept the responsibility for carrying out the observation to local standards

  • all staff must receive appropriate training in this role especially staff for whom this role is new.


7.1 As the only profession which has 24-hour contact with patients, nurses are in a key position to ensure robust safety and therapeutic care. As stated, it is still nurses who will remain the staff group predominantly involved in observation and their experience in this skilled task must be recognised and utilised by the other professions. The nurse in charge of the ward or unit should retain responsibility for co-ordination of decisions regarding observation levels. For clarity only one individual must be charged with this duty and be seen as the person who should always know both who is being observed (at a raised level) and which staff member is responsible for a given time period. That person is the nurse in charge.


8.1 Caring for someone in distress is not a process to which only professionals can contribute. Anyone with a suitable approach and awareness can, at times, help and may, on occasions, be more appropriate than the professional. Clearly it is neither safe nor fair to expect a carer or fellow patient to shoulder the full responsibility of caring for someone in severe distress. However, in many settings the patient being observed may welcome the company of a relative, friend or fellow patient and it may not be appropriate to have a staff member present during such occasions.

8.2 During general observation there would appear to be no conflict, whereas during special observation it would not be appropriate to leave a patient without a member of staff present. However, constant observation is more complex. There will be situations where it is reasonable and appropriate, and others where it would be unsafe and unfair. The risk assessment process and subsequent multi-disciplinary team discussion must include decision making; agreement should be reached on the appropriate level of observation and who can offer the greatest level of support to the patient.


9.1 Patients and their carers/relatives should be informed of the observation policies and procedures in use within the service. If observation is to be a true partnership then clear, honest and open dialogue must take place regarding the reasons for an increased level of observation. Written information regarding observation policies and practice must be given to all patients. Specific information regarding a patient's current level of observation must also be given. It is recommended that this information is developed in conjunction with local service user groups.

9.2 Although it is not appropriate for a formal "appeal" (as in the Mental Health Act) patients should have access to an advocacy service to assist and guide them in disagreements about any restrictions on their freedom that observation may lead to. Written explanations should be given to both patient and carer/relative about the level of observation in use and its purpose. Patients should also be offered an opportunity to discuss their concerns with a senior member of staff.

9.3 If the process is designed to be truly collaborative and crucially "feels" this way to the patient then the chances of the patient disagreeing with the decision are reduced. Patients must have the right to discuss formally their views on their observation level with staff and, if they desire, involve someone (such as from an advocacy service or friend/relative) in these discussions. To facilitate this collaborative process local user groups should be encouraged to become involved in the development of local observation policy, written information and staff training.


10.1 The need for high-quality training was highlighted in the original document. It would appear from our review that training in the practice of observation is still seen by many as an issue requiring more attention. The specific need for training non-nurses in observation responsibilities has already been referred to in this document. It should be noted that although nurses are traditionally the profession most closely associated with observation it would appear that few have received specific training. The review group was not aware of any current formal courses but recommend that local services develop training plans for all staff involved in observation. This training should include input from users and should explore both the practicalities of the local observation procedure and the philosophy underpinning it.

10.2 The training must also include clinical skill training as needed to enable staff to have the necessary "tool-box" of psychological and practical interventions to help patients cope with their distress and illness. Skills such as distress tolerance, suicide thought reduction, psychotic thought management, problem solving and anxiety management are appropriate to include. These two elements of the training should be aimed at ensuring the process of observation is therapeutic and safe.

10.3 The recent Leadership Development Programme is one route through which to identify key players in implementation of local observation protocols. Reference should be made to "Learning Together _ A Strategy for Training and Lifelong Learning" (1999).

10.4 Training must not be seen as a "one-off" as, apart from ongoing updates, all staff working in psychiatric units need continuing organisational support in their clinical work. This may be through a system of clinical supervision, mentorship or preceptorship. The recent Nursing Strategy "Caring for Scotland _ The Strategy for Nursing and Midwifery in Scotland" (2001) emphasises the important role of clinical supervision in nursing. The need for clinical supervision applies to all professional groups and must be seen as an integral part of good quality care and not an optional extra. It is also part of supporting the process of learning or reflective practice that comes from both formal audit and critical incident reviews and contributes to the continuing professional development that is essential to all professionals.


11.1 Much can be learned through careful analysis by the multi-disciplinary team of the management of particular incidents which should not have occurred. Critical Incident Reviews (CIRs) are considered to be a valuable learning tool for staff, as well as a supportive forum. All wards/units caring for acutely-ill patients should adopt the practice of carrying out reviews when untoward incidents occur (the "critical" applies to the seriousness of the event for the organisation's overall purpose and not to an expressed intent to pin the "blame" on somebody). Although CIRs are associated with the investigation of suicides, suicide attempts and incidents of violence and aggression, the process is mentioned here as a means of looking at "near-misses". By these are meant any failure of systems or failures to apply the correct procedures. Therefore, when an event occurs which constant or special observation was designed to avert,
it is worth examining to see how in future the team (and the organisation) can ensure a better outcome. A protocol for conduct of the CIR is described in "Risk Management". To ensure they are useful learning tools, CIRs should be conducted in a learning mode, separate from and not linked to disciplinary procedures, and should be carried out in a manner that facilitates development both on a personal and organisational level.

11.2 If CIRs occur frequently then there may be something wrong with the organisation's ability to learn and incorporate lessons into its risk management processes. More can be gained from looking honestly at occasions when things did not go well, but there is a delicate balance of formal and informal systems applied in a human context. This issue is explored within "An Organisation with a Memory _ Report of an expert group on learning from adverse events in the NHS" (Department of Health 2000) and the follow-up document "Building a Safer NHS for Patients" (Department of Health 2001).


12.1 Clinical Governance is defined as "corporate accountability for clinical performance". A key purpose of Clinical Governance is to improve quality of care and to ensure that wherever possible poor performance is identified and addressed. Locally, Trust Clinical Governance Committees are expected to ensure that their organisations put in place systems to allow for learning from complaints or critical incidents. It is essential that observation practice, along with the risk management inherent within it, is seen as a key mental health issue for examination through the Clinical Governance process.

12.2 Simple auditing of the frequency and duration of observation practice across wards is an essential tool in monitoring the effectiveness and usage of this procedure. For an example of this approach see Porter, McCann & Kettles (1998). The audit process must also ensure that feedback from users is gathered and used to shape both policy and training.

12.3 Recording systems must be designed to allow a clear audit trail. The record must be simple and quick to complete but should include: the level of observation used, the presenting clinical picture, who is involved in these decisions, the views of the patient and carer, information given to the patient and carer, and any specific plans for possible reduction of level of observation. Pending the development of electronic recording of practice data, a simple paper record will suffice. Such records are essential components in the discussions at Critical Incident Reviews and in the defence of professional practice in the event of a Fatal Accident Inquiry.


13.1 Caring for people experiencing periods of acute mental illness or distress is challenging and demanding _ experiencing it must be even more difficult. Comprehensive care involves both ensuring the patient's safety and offering therapeutic help and support. One without the other is incomplete. The risk is that services pay considerable attention to the safety component but perhaps less to making the observation experience healing and compassionate. To ensure both are achieved staff must feel valued and supported and given the time and skills to truly engage with the patient. The process must be a partnership approach between staff and patient where the needs of the patient are recognised and respected. There is much one group may learn from the other in how safe, therapeutic observation can be developed and delivered.

Back to top