The Use of Control Interventions in Adult Mental Health Inpatient Services in Ontario Hospitals and Identification of Risk Factors

Abstract

The use of control interventions (CI) such as mechanical/physical restraints (MP), chairs that prevent rising (Chair), and acute control medications (ACM) with mental health (MH) patients is not without controversy. Clinicians report using CIs with patients to gain immediate control of a situation of harm or imminent risk of harm involving the patient and/or others and not unexpectedly, there are unintended physical, psychological, emotional injuries and in the most serious of unintended consequences, death. Despite these substantial negative consequences, there is a dearth of research on the use of CIs in MH hospital services.

The goals of the current study were three-fold: to establish the prevalence of CI use and to profile the sociodemographic, MH service use, and MH clinical characteristics of adult MH inpatients with CI use; to better understand the potential risk factors for the use of MP, Chair, ACM, and an inclusive category of “any control intervention”; and thirdly to examine the use of CIs with adult MH inpatients in the absence of a psychiatric emergency situation (NoPES).

Methods: The study sample included adult inpatient mental health patients with an admission Resident Assessment Instrument- Mental Health assessment between 2006 and 2010 in Ontario. A descriptive analysis was conducted and multivariate logistic regression was used to investigate potential risk factors for the use of control interventions and the same investigation was conducted for a subsample comprised of patients presenting in hospital without a psychiatric emergency situation. Control interventions included mechanical/physical restraints, chairs that prevent rising and acute control medications.

Findings: In Ontario, 21.0% of MH adult inpatients had at least one kind of CI use in the study sample (N = 115,384). The most frequently used CI type was ACM (18.6%) followed by MP (6.5%), and Chair (0.9%). The risk models for each CI type was more informative than for an all-inclusive category of Any-CI. Aggressive behaviour, mania, positive signs and symptoms, risk of harm to others and severity of self-harm increased the risks for CI use. Non-voluntary admission increased the risk of use for all CI types. Unexpectedly greater deficits in performing of activities of daily living, instrumental activities of daily living, cognitive performance, self-care, and having a history of falls increased the risk of CI use. A focus on these latter risk factors will be important in the development of CI reduction strategies. Sociodemographic and history of health service use variables were also considered in the model including older age which was uniquely a risk for Chair use whereas younger age was a risk for MP and ACM showing a bias or preference for Chair use with older patients. Gender was a significant risk factor for MP, ACM use, and only for Chair use with NoPES. Seventy four percent (or 85,514) of the sample did not present in the three days prior to assessment with a psychiatric emergency situation (NoPES) such as extreme behaviours that may put the patient or others at risk of harm. Of these NoPES patients 12,097 (14%) experienced CI use regardless. The identification of NoPES patients is an immediate opportunity for reducing the use of control interventions in MH. The risk models developed in the current research can inform the development of CI reduction strategies; deficits in functional performance increased the risk of CI use as did a history of falls, and attenuated aggressive behaviour, mania, positive signs and symptoms and risk of harm to others. If CI use could be eliminated for the NoPES patients, Ontario could reduce its CI use to less than 10% achieving a major step to providing higher quality patient care for patients and increasing staff satisfaction with the care.

Conclusion: This study made use of the new interRAI Control Intervention clinical assessment protocol (CAP) which is embedded in the RAI-MH tool to identify the NoPES patients. This innovation in the RAI-MH is readily available to all hospitals currently using the hospital-based MH assessment instruments. Ontario is in a unique position to immediately use this capability to advance a quality improvement initiative to reduce the use of CIs in MH. The data is readily available enabling public reporting and benchmarking on CI use rates as a patient safety indicator as well as providing hospital-level reports. Additionally, more study is needed nationally and internationally to increase our knowledge of why CIs are used in MH and consequently create effective staff education/training strategies to reduce their use. In conclusion, the use of CIs in adult MH inpatient services in Ontario requires the attention of policy makers and hospital administrators. If CI use in adult MH inpatient services was identified as a priority patient safety concern by government (as it has done for hand hygiene, hospital mortality, and medication safety), Ontario could use readily available data on CI use to immediately measure prevalence, establish performance targets, and report on the progress of improving the quality and safety of care of adult MH inpatients.