To the Editor: Millar and colleagues recently described their comparison of the national inpatient medication chart (NIMC) with 14 other medication charts.1 They concluded that the NIMC contained design features that were adverse and therefore inferior to the medication chart previously used in their hospital. They also stated that the advantages expected by the Western Australian Director-General of Health in introducing the national chart were not experienced at their hospital.

Millar et al failed to mention that the NIMC underwent an extensive process of piloting and evaluation in over 30 sites across the country in a structured before-and-after study.2 Failure to recognise (i) the benefits of standardisation as medical, nursing and pharmacy staff move between sites, (ii) the opportunities for structured safe medication practice training,3 and (iii) the value of the collaborative methods used will inhibit the possibility of overcoming problems like those identified by Millar et al in future redesign processes. Millar and colleagues themselves noted that “marked heterogeneity of chart design has been abolished by the NIMC”.

The national pilot study considered the entire medication management cycle using a broad definition of medication error (“A prescribing decision or prescription writing process that results in an unintentional, significant reduction in the probability of treatment being timely and effective or increases the risk of harm, when compared with generally accepted practice”4). The NIMC was designed to reduce the risk of errors that prescribers have identified with previous charts.5 The NIMC also reduced the need for all staff to interpret unclear or incomplete prescriptions, thereby further reducing the risk of medication errors.2

We support the comments by Millar and colleagues that the process of implementing clinical practice change must involve significant buy-in and championing by clinicians. The implementation of the NIMC in Queensland recognised the importance of top-down support from senior health officials, combined with the need to increase clinicians’ awareness of risks of current systems and the need for a clear demonstration of the benefits of a revised system to bring about any substantial change in behaviour.

We understand that the Australian Commission on Safety and Quality in Health Care has established a quality assurance process which operates at jurisdictional and national levels to adjust the NIMC on the basis of issues raised. This important platform will succeed in addressing the issues raised by Millar et al provided clinicians participate in this collaborative approach to medication safety. We have a rare opportunity, in which Australia is taking a leading role, to address one of the critical safety risks facing patients today. Let us all work together and criticise constructively within a framework of collaboration.

In reply: It is understandable that the designers of the national inpatient medication chart (NIMC) should wish to defend it against criticism, especially after 5 or more years of hard work and the major administrative achievement represented by the “top-down” implementation. It is regrettable that the chart at the centre of this otherwise admirable activity turns out to have significant weaknesses compared with the previous medication chart used at Royal Perth Hospital, and that the designers acknowledge this only obliquely by allowing for “future redesign”. Rather, they emphasise secondary outputs such as cross-border familiarity (which we discussed in our article1), “training in structured safe medication practice”, and “collaborative methods”. These supposed advantages are but small crumbs of comfort compared with the imposition of an unsatisfactory chart, loss of local autonomy and increased hazard for patients. There is no evidence that the NIMC has decreased medication errors, defined in relation to patient harm. There was indeed a pilot study, and we referred to it in two different contexts in our paper, but it assessed the chart on the basis of unsatisfactory process-based criteria similar to those employed after the chart was implemented. Perhaps a better indication of the problems of the pilot chart lies in the hundreds of suggested changes made from pilot sites to the NIMC Oversight Committee.2

We note that Coombes and colleagues do not dispute our scientific findings or the design faults we described. Their response repeats unverified claims of benefit that we discussed in our article, and seeks to reassure readers that a process is in place to “adjust the NIMC on the basis of issues raised”, thus acknowledging that “issues” exist. However, readers should be aware that the process referred to is subject to a set of ground-rules which prohibit changes to several design aspects of the chart that we criticised (eg, the block design of the pro re nata [PRN] section).3 Thus, the possibility that the NIMC will be substantially improved is remote. A more likely outcome is that Australia will be left with a chart that satisfies the superficial attraction of national standardisation but contains significant design flaws which represent a hazard to patients. A better approach would be to agree on binding national standard design elements and to restore to individual hospitals or health areas the right to design their own charts within these constraints — “think globally, act locally”.4