Standards,efficiency and effectiveness in consultation-liaison psychiatry

Given growing clinical, economic and systemic arguments, addressing the mental health needs of patients with physical presentations should be integral to normal service delivery. Consultation-liaison psychiatry (CLP) is a subspecialty which interfaces between general medical and psychiatric services. The variability between models of CLP services has rendered direct comparisons difficult, and a lack of consistency in evaluation studies reflects this (Wood and Wand, 2014). This paper presents recent developments in measuring standards, performance and outcomes in CLP, and suggests directions for further work.

Standards of accreditation and training

In Australia and New Zealand there is no consensus on what CLP services should look like and which standards should be met. The UK Royal College of Psychiatrists (RCP) established the Psychiatric Liaison Accreditation Network (PLAN) as a quality improvement initiative (http://www.rcpsych.ac.uk/workinpsychiatry/qualityimprovement/qualityandaccreditation/liaisonpsychiatry/plan.aspx). The standards set out in PLAN include service structure, service delivery and procedures, evaluation of quality and scope of practice. An additional strength is the peer review inspection visit, which includes service user ‘experts’. There are also dedicated standards for working with the Emergency Department, urgent referrals and for various age groups.

It is not only health services, but training programs for specialists which are under scrutiny. Internationally there is a focus upon competency based training in psychiatry. Specific competencies are defined for CLP, but are limited. The UK Shape of Training review emphasised the need for more general medical training for trainee psychiatrists to better address the needs of patients, recognising that increasingly they are older and have multiple comorbidities. This has led to calls for better training in physical health in psychiatry, including medical placements in neurology and related subspecialties like geriatrics.

Reviews of CLP services to older adults in the UK have also revealed heterogeneity in service models and no agreed standards or guidance for data collection, service provision and ownership ¹ . It has been argued that specialist dedicated CLP services to older adults which are hospital-based are more efficient, improve pathways of care and reduce length of stay and readmission ¹ (Tadros et al., 2013). There is no current requirement in the UK, Australia or New Zealand for trainee psychiatrists in CLP to undertake placements in psychogeriatrics ¹ .

Measurement of efficiency and performance

A measurement framework for CLP was recently proposed by Fossey and Parsonage (2014). The main recommendations were the adoption of a systematic model where a range of measures are linked to the objectives of service provision. Three aspects of service delivery should be assessed; structure, process and outcome. Structure refers to the available resources in the setting of the CLP service; process to the specific activities of the service in delivering healthcare; and outcome to the results of this activity.

Surprisingly, there is scant international discourse about the ideal structure of CLP services. A recent Guidance from the UK outlined four grades of service models for CLP ranging from minimum (Core) to optimum (Comprehensive), which are considered cost effective and to have quality benefit (Aitken et al., 2014). The detailed ideal structure of CLP services is well described by the Guidance and PLAN, including staff mix and specialist input and minimum numbers for hospital size taking into consideration locally expected roles of the CLP service.

Process measures in CLP have been directly measured, including response times, the number, frequency and timing of referrals and health utilisation data such as types of admissions, discharge destination and readmission rates. There are accepted benchmarks for timeliness of response to defined categories of referrals (emergency, urgent, routine). A recent study has demonstrated reduced length of stay) and lower readmission rates for CLP-referred patients and greater return to community living in older adults (Tadros et al., 2013).

Measurement of effectiveness and outcomes

In an increasingly economically driven healthcare system there is pressure to demonstrate transparency, cost effectiveness and improved outcomes for patients and the health service (Fossey and Parsonage, 2014). Indeed, CLP services in general hospitals are cost effective and reduce length of stay, when involved early in admission (Wood and Wand 2014). Unsurprisingly, given that CLP services have different roles, settings and structures, there is no universally accepted set of measures of effectiveness and outcomes (Fossey and Parsonage, 2014; Wood and Wand, 2014). Part of the problem is that CLP services vary considerably in their activities, many of which do not fit neatly within the typical functions of general mental health or physical health services.

Another aspect to be considered when evaluating outcome is the CLP role of up-skilling of general hospital staff. There is recent evidence from a specific CLP model that regular support and mental health training of acute hospital staff contributed significantly to cost savings through reduced length of stay and readmission rates in patients not directly referred to CLP (Tadros et al., 2013). Similarly, there is evidence that multicomponent educational interventions from CLP targeting delirium risk factors (using lectures, case-based tutorials, delirium resource staff and feedback) reduce the incidence of delirium (Wand et al., 2014).

It is not meaningful to simply transpose the outcome measures used by mental health services, such as the Health of the Nation Outcome Scales (HoNOS), to CLP. CLP services may only see a patient once, limiting the utility of before and after outcome measures, and the focus of intervention may not be reducing symptoms.

Consumer experiences, including family/carers, are important barometers to inform improvement. Most tools measuring consumer satisfaction in CLP are locally derived, unvalidated and not repeated, making comparisons between services difficult (Wood and Wand, 2014). Patient Reported Outcome Measures (PROMs) and Patient Reported Experience Measures (PREMs), used in acute trusts in the UK, have been proposed as superior to satisfaction questionnaires because they comprehensively canvas the consumer perspective.

A broad range of outcome measures have been proposed for CLP, with the main dimensions being clinical improvement, satisfaction (of both patient and referrer) and impact upon service use (cost effectiveness, length of stay, readmission rates) (Fossey and Parsonage, 2014; Wood and Wand, 2014). Although PLAN requires that a CLP service regularly review themselves through audit, quality and feedback for accreditation, they do not specify which tools should be used. It is not sufficient to consider only one of these dimensions when assessing effectiveness in CLP, as the scope of practice is broad and interfaces between mental and physical health (Wood and Wand, 2014). Further, the complexity, multi-morbidity, and variable presentations of patients seen by CLP make it difficult to attribute clinical improvement to the CLP intervention alone.

Conclusions

Peak international bodies for CLP should combine expertise and perspectives in a working group to develop universal (minimum) standards for CLP services. Then, consensus is needed upon a set of evidence-based measures of efficiency and effectiveness; the latter encompassing a range of objective and subjective CLP outcomes. Older inpatients must be a focus and experience in psychogeriatrics mandatory for CLP trainees. One approach would be to use paired measurements at the start and finish of CLP involvement with patients to evaluate individual benefits attributed to CLP involvement and guide quality improvement. A universally applicable consumer evaluation tool is urgently needed as none exist for CLP services.

Further research will help determine the optimal way of measuring processes and outcomes in CLP. A multicentre case-control study may be the most appropriate methodology to evaluate the implementation and viability of such measures. This is an exciting time for CLP clinicians internationally to unite and demonstrate the value of their multifaceted input with patients, general hospital staff and health service systems.