Williamson's commentary (2007) poses useful questions and insights on our research and offers readers additional perspectives regarding the use of participatory action research to improve clinical practice. The authors are pleased that the paper has invited critique and the opportunity for a response. It is salient to point out that this paper reports on a Masters of Nursing project. Therefore, the scope and timeframe of the study was limited to that comparable with a two-year project. Questions about the limited scope of the study, such as the level of participation of patients and family and the choice to exclude staff working outside of the acute care sector, are relevant to a larger study. For example, Williamson suggests this study could have been strengthened if patients had continued to be involved in other stages of the research process. Examples include contributing to the selection of a care model that would best meet their needs, participating in the procedures involving the evaluation processes and in determining patient-focused outcomes. Apart from time constraints for the Masters student, this aspect may have been challenging due to the acute medical ward environment, age, medical condition and knowledge levels of patients; and the time constraints, work load and knowledge levels of clinical nurses participating. Patients, family members, carers and nurses were involved in evaluating care received and given during the premodel and patients and nurses during the model as well as the implementation stage.

A further point made by Williamson is that she asks why particular models of care were presented as options and why a choice was made from existing models as opposed to developing one. Although many Action Research studies have been successful in enabling nursing staff to identify, develop and monitor a care model of their choice, this was not possible in this study. In this project, there was a 12-month time limit on the development and monitoring of practice change, dictated by the maximum two-year timeframe allowed for the Masters degree candidature. The nurses on this particular acute medical ward wanted the researcher to give them a variety of models to choose from, because of their own time restraints. It was difficult enough for nursing staff to find time to attend the meetings with the research team as it was, let alone the time it would have taken for them to research all the models of care without prework done by the researchers. The nurses were offered a range of care models proven to be effective in meeting older patients’ care needs and they were encouraged to select one, that was best suited to their needs, and to tailor it to the specific needs of the ward. Further information on this process can be found in the unpublished thesis by Glasson (2004).

The commentary recognizes the patients’ knowledge levels of their medication regimes were improved following repeated educational sessions during the model of care over the period of their acute hospital stay as indicated by the study. The commentary also suggests the transfer of older patients between settings, such as when discharging from hospital to home, presents particular danger points for medication errors to be made (Banning 2005). Williamson makes a valid point that it would have been useful to involve caregivers in the education sessions; however, many patients did not have caregivers and it would be feasibly difficult to arrange caregiver presence at the same time medications were administered. Nonetheless, this should be considered in future research, bearing in mind that potential subject numbers would be limited. Our study recommended that future models should focus on multidisciplinary approaches involving physicians and allied health professionals to monitor older patients’ knowledge levels of their medication regime after discharge from the acute hospital setting. This was based on the recognition that nurses do not deliver care in isolation, and particularly in the case of older patients, a multidisciplinary team approach is applied.

Williamson also questions why only registered nurses were involved in medication education sessions and what impact those nurses who did not want to take part had on the model. In Australia, only registered nurses are legally qualified to administer medications in the acute care sector and the majority, if not all, nurses in the acute care sector are registered nurses. As the education sessions occurred during the administration of medications, only registered nurses were present. Secondly, those nurses not involved in the model of care study (mainly casual or agency nurses), were informed of the study and their patients (if eligible to participate according to the selection criteria) were recruited by permanent nursing staff participating in the model of care study. It should be noted that the majority of nursing staff were involved in the action research implementation.

Another point made in the commentary implies there may be a mismatch between what the patient needs, and what they want. For example, although the patient may recognize what would be good for them (self care), this does not necessarily match how they would prefer things to be (have things done for them by nurses). In this model of care the nurses were aware of these differences and addressed this concern by always adding a rationale to their request for them to attend to their self care. For example, when asking the patient to walk slowly to the bathroom to attend to their daily shower or walk around the ward corridor (when appropriate to the condition), they were informed of the benefits of this activity. For example, this activity may improve their blood circulation to help prevent their blood forming blood clots, and may improve their mobility.

Regarding the environmental setting from a leadership style viewpoint, further details can be found in Glasson's thesis (2004). Briefly, leadership style in this acute medical ward was less participatory (nursing staff take part in problem-solving and decision-making) and more bureaucratic (supported by procedure and policy manuals), which are typical of many traditional hospital approaches (Speedy 2004). The staffing levels on this ward were stable during the premodel and model of care study, which we believe gave the model the best chance of being implemented and changes sustained. The ward was managed by two nursing unit managers who job-shared, both of whom encouraged nursing staff to take ownership in the development and implementation of the model.

The authors would like to thank Williamson for her commentary and for congratulating us for our efforts in including older participants with co-morbidities. Her comments are valuable, especially regarding how to maximize participant involvement in the participatory action research process. Her feedback to the authors is much appreciated.