“Not-for-resuscitation” orders in Australian public hospitals: policies, standardised order forms and patient information leaflets

“Not-for-resuscitation” (NFR) orders are a necessary aspect of current medical practice in Australia.1 They are implemented to prevent the use of cardiopulmonary resuscitation (CPR) in situations when it is deemed futile or unwanted. The term CPR refers to a range of resuscitative efforts, including basic and advanced cardiac life support to reverse a cardiac or pulmonary arrest.

NFR ordering procedures vary considerably between institutions worldwide.2-4 Similarly, there is no universal term for NFR, with terms such as “do not resuscitate” (DNR), “do not attempt resuscitation” (DNAR) and “not for CPR” also used.

A standardised order form is a pre-printed form used to document an NFR order, sometimes complementing clinical note entries. Patient information leaflets on NFR decision making have been made available to patients and families. Various Australian and British health bodies recently recommended the implementation of hospital NFR policies, standardised order forms and patient information leaflets, also providing a framework for policy content.1,5-7

The prevalence of these documents in Australia and their content have not been evaluated in detail (as shown by a search of PubMed and Ovid MEDLINE databases, using the search terms “resuscitation orders”, “policies”, “prevalence”, and “content”, from 1966 to December 2005).

The purpose of this study was to determine the prevalence and analyse the content of hospital policies, standardised order forms and patient information leaflets for the use of NFR orders in Australia.

All Australian public hospitals with 60 or more beds were included in the study. Hospital details were obtained from the Australian Hospitals Directory.8 Private, psychiatric and military hospitals were excluded. Hospitals were stratified by bed numbers, academic status and location.

The Director of Medical Services of each hospital was mailed a letter explaining the purpose of the study and including a one-page questionnaire, asking if they had policies or guidelines, a standardised order form or patient information leaflets pertaining to NFR orders used at an institutional level. If so, they were asked to provide copies of the documents.

Hospitals were given the option of responding via mail or email. If the hospital did not have a Director of Medical Services, then the letter was addressed to the Director of Clinical or Nursing Services. Hospitals that did not respond within 6 weeks were contacted again by mail and provided with the extra option of faxing their replies. Data were collected between August and December 2005.

Of the 222 hospitals that satisfied the inclusion criteria, 157 responded (71%). There was no significant difference in the response rate based on hospital size, academic status or location. Hospitals that indicated they had relevant documents but did not provide them or provided documents that did not meet the criteria for an NFR policy, SOF or PIL were categorised as “responding hospitals without policies/forms”.

Just over half the hospitals surveyed (54%) had NFR policies, in line with other developed countries.2,9 Fewer hospitals (39%) used SOFs, and only four of the 157 respondents (3%) had PILs that we could analyse. If hospitals that were in the process of formulating the above documents were included as already using them, then the prevalence would increase to 68% for NFR policies, 49% for SOFs and 10% for PILs.

Larger and metropolitan hospitals were more likely to have NFR policies, possibly reflecting their larger budgets and greater awareness of the need for such policies. Introduction of an institutional NFR policy results in increased patient involvement,12 improved documentation13 and comprehension12 of NFR orders, and fewer inappropriate resuscitative efforts in terminally ill patients.14,15

Almost half the policies analysed (47%) did not define an NFR order. This is a concern, as it has been shown that Australian doctors and nurses differ in their perception of NFR orders.1 Furthermore, 38% of policies did not indicate the person responsible for documenting orders. More policies should provide a proper definition and clear delegation of responsibility to avoid confusion between medical and nursing staff.

While a high proportion of policies did not address the issue of disagreements within the medical team (79%) or between the medical team and the patient or family (60%), these figures are lower than those in comparable studies of Dutch and Canadian hospitals.2,9 Existing policies should be reviewed to outline mechanisms for conflict resolution.

It is alarming that 62% of NFR policies did not address the issue of advanced care directives or “living wills”, which have been shown to affect treatment decisions.16 The issue here is not whether they should be accepted as legal documents (which may vary from state to state, depending on legislation), but the fact that a large proportion of policies do not mention the issue. Hospitals should update existing NFR policies to reflect the current legal situation in their respective state or territory.

Most policies (79%) did not outline any criteria for labelling a patient as incompetent. Possibly, the criteria are listed in another hospital protocol, but none of these policies referred to another such document. Only 34% of policies stated that a new NFR order was needed for a patient being re-admitted to hospital who had an order from a previous admission. While this figure may be higher in practice, more policies should indicate that a new order should be sought for every admission, as the patient’s health status and preference may have changed. However, it should be noted that no policy indicated that an order from a previous admission would be valid.

Among the strong points of Australian NFR policies is the fact that 89% identified the members of the medical team allowed to make NFR orders, compared with 75% of Dutch policies.2 Also, 86% of Australian policies stated the need for NFR orders to be reviewed, which mirrors the proportion in Canada (86%)9 but is higher than that in the Netherlands (61%).2 Regular review of NFR orders is important because of the possibility of significant clinical improvement that may result in resuscitation no longer being considered futile. However, 58% of policies did not outline the procedures to reverse NFR orders. A high proportion of policies in Australia (89%) indicated the need for a competent patient to be involved during the process of discussing NFR orders. Similar figures are reported elsewhere,2,9 indicating that there is global awareness about the importance of patient communication when implementing these decisions.

We found wide variation in the content of SOFs for the use of NFR orders. One could argue that particular details in NFR orders not included in the SOF could have been written in the patient’s clinical notes. However, only 13% of NFR policies indicated that NFR orders were to be written in both the SOF and clinical notes. A reassuring point of Australian SOFs is that 81% included documentation of discussion with the patient. Studies analysing use of SOFs show that NFR decisions are significantly more likely to be authorised by consultants,17,18 be in accordance with hospital guidelines,17 be reviewed when necessary,17 show clarity of documentation,18 have active patient participation17 and document reasons for the decision.18 More hospitals should therefore adopt the use of SOFs, from the current 39%.

Analysis of the content of PILs about NFR orders is limited by their low prevalence (3%) in Australian hospitals. PILs have been rated as useful and not a cause of unnecessary distress or offence.19

Our study had several limitations. Content analysis is potentially open to subjective bias. To reduce this likelihood, all documents were analysed by two independent reviewers, with differences resolved by consensus or a third reviewer. Hospitals with fewer than 60 beds, most of which were small district hospitals and respite care centres, were not included in our study as we considered that our preformed structured list of items would not apply to most of these if they had the relevant documents. A higher response rate could possibly have been obtained by including a self-addressed stamped envelope with the survey.

A strength of our study was that there was no significant difference in the response rates based on hospital size, academic status or rural versus metropolitan location. Our findings are therefore less likely to be susceptible to non-response bias.

In conclusion, an NFR policy should be clear, concise and relevant to the hospital setting. Nevertheless, the presence of a policy does not guarantee that it will be followed.20 Proper understanding and implementation of an NFR policy may not be possible if hospital staff are not aware of its contents or if the document is seen to be too complicated.21 Australian hospitals should be encouraged to formulate NFR policies, utilise standardised order forms when implementing these policies and allow for the provision of patient information leaflets. The content of existing documents should be reviewed based on the findings of this study, which add an Australian perspective to an important issue. Variations in state/territory recommendations and legislation should be taken into account when formulating these documents.