Abstract
DEAR SIR,
Dr Ratcliff states that the primary purpose of the clinical record is ‘to preserve clinical information and provide a basis for judging the progress of the case’. Unfortunately, the conventional medical record often preserves clinical information in the same way that volcanic ash preserves Pompeii: in faithful outline, but requiring great archaeological skill to uncover.
An effective record should not just preserve. It should provide a structure for recording, prioritizing and retrieving information for multiple clinicians over different settings and times. The conventional medical record often does this so poorly that our choice is not between a poor form and an acceptable status quo. Our records are broken, and need fixing. The current NSW Health Mental Health Outcome and Assessment Tools (MHOAT) forms have been an imperfect attempt to address these concerns.
Most public mental health clinicians regularly experience the limitations of the conventional clinical record, including the following.
The frustration of patients and carers at being asked repetitive biographical or historical information by successive interviewers.
The struggle to make sense of a complex assessment with a missing or incomplete file.
The sinking heart associated with searching multiple volumes for a succinct summary of treatment history or the reasons for previous clinical decisions.
The challenge of trying to trace the original source of information regarding serious risk back through its successive recounts and summaries.
The difficulty of searching for a calibrated record of past clinical progress against which to judge current treatment (and the despair at finding a long and complex admission summarized as ‘Settled well in hospital’).
These everyday failings of clinical documentation lead to poor outcomes, or to missed opportunities for good outcomes. They reflect intrinsic structural limitations of the medical record that are increasingly recognized throughout health – the cardiologist Don Berwick has quipped that the conventional medical record ‘is not medical and is not a record: it's a novel about a fever’ and argued that design of more effective records should be health's current ‘moonshot’ project.[1]
Documentation of risk assessment provides an important example. Mental health clinicians balance complex judgement, inexact science and public expectation on a daily basis. We usually succeed, but our records often are a poor reflection of our efforts. A recent review of 120 files of suspected suicide deaths within 28 days of discharge from inpatient mental health care examined for any documentation of suicide risk in either structured risk assessment forms or free-form clinical notes.[2] Even adopting a low threshold, the assessments of 43% of patients admitted following a suicide attempt contained no discussion of suicide risk. Where patients were reviewed by a medical offi-cer before discharge, more than half of the entries included no assessment or discussion of risk. Less than one-fifth of discharge summaries included any discussion of risk. These are not isolated findings. They reflect a failure of the medical record as a tool for communication of important clinical information at critical points in the process or handover of care.
Structured medical records are an increasingly frequent response to such documentation problems.[3] Approaches range from paper pro formas to computerized health records and decision support systems. Such records are no panacea: many have failed through poor design and implementation. Kawamoto et al., reviewing randomized trials of clinical decision support systems, suggest that successful systems focus on ease of clinician use and speed of feedback, and are usually computerized.[4]
The Ethical Practice Committee articulates important principles for improving our records. Their position may be summarized as follows.
The use of structured clinical records is not intrinsically unethical.
Poorly designed or excessively lengthy forms may create problems through double-handling of information, distortion of the clinical process, loss of rapport and ineffective use of clinicians' time.
Collection of secondary information not relevant or appropriate to the clinical encounter should be of low priority for clinicians.
Documentation of risk assessment should reflect its dynamic nature and the need for constant and careful review.
Both the conventional medical record and the current NSW MHOAT forms suffer frommany of these problems. On some issues and in some settings, the forms have probably improved on the conventional record. On other issues, they may have inadvertently reduced the quality of clinical documentation.
The NSW MHOAT forms will be evaluated, and substantially revised. This process will be managed by Infor MH, a unit newly established to provide information for mental health service evaluation and development in NSW. We are keen to add some evidence to the debate, and are planning to commence an evaluation of the performance of the current forms and conventional records against clinically derived standards. We would welcome feedback on weaknesses or strengths of the current forms, or examples of effective structured or semistructured clinical record systems. We welcome the Ethical Practice Committee's contribution to the challenging task ahead.