The use of clinical practice guidelines for community-acquired pneumonia in hospitals in Atlantic Canada

Department of Emergency Medicine, Dalhousie University, Halifax, Nova Scotia Correspondence: Dr Samuel G Campbell, Department of Emergency Medicine, Queen Elizabeth II Health Sciences Centre, 1796 Summer Street – HI 3021, Halifax, Nova Scotia B3H 3A7. Telephone 902-473-3871, fax 902-473-3617, e-mail samuel.campbell@cdha.nshealth.ca SG Campbell. The use of clinical practice guidelines for community-acquired pneumonia in hospitals in Atlantic Canada. Can Respir J 2004;11(4):301-303.

C ommunity-acquired pneumonia (CAP) is a common afflic- tion, and its treatment contributes to a significant portion of health care resource use (1).Considerable variations in treatment have been identified, both in the antimicrobial agents selected and in the thresholds for admission and discharge from hospital (2)(3)(4)(5).A number of process variables have been strongly related to outcome (6,7).Clinical practice guidelines (CPGs) have been shown to improve the care of patients with CAP, and have resulted in fewer hospital admissions, shorter hospital stays, shorter periods on intravenous antibiotics and improvement in the aforementioned process variables, such as time to administration of antibiotic and antimicrobial selection (6,(8)(9)(10)(11).CPGs are considered to be the standard of care by most authorities in the treatment of CAP, yet it appears that CPGs have yet to be adopted in many hospitals.Knowledge of the extent to which CPGs have been adopted is important because it can direct where efforts can be expended to improve the care of CAP patients in Atlantic Canada.

METHODS
A telephone survey was conducted of all hospitals listed in the Canadian Medical Directory as being situated in Atlantic Canada.Surveyors requested for interview were, in the following order, the head nurse, the chief of medical staff or the hospital pharmacist.Hospitals were excluded from further analysis if they did not have all three of the following: an emergency department, x-ray facilities and acute care inpatient beds.A structured survey (Appendix 1) was used with a standard set of definitions for each term used (Appendix 2).Answers were entered directly onto a computerized database developed by the database specialist of the Dalhousie University Department of Emergency Medicine (Halifax, Nova Scotia), and data were analyzed using Microsoft Access (Microsoft Corporation, USA).
province is shown in Table 2. Of the total number of hospitals, 44.1% reported using CAP CPGs, 49.5% did not and 6.5% did not know whether they did.Hospitals with more than 300 beds (n=3) (Table 2) were universally using CAP CPGs, with the proportion of users declining steadily with the number of beds.Urban teaching hospitals and those with an intensive care unit were significantly more likely to be using CAP CPGs (Table 2).Forty-four per cent of hospitals using CAP CPGs reported that their program had been in place for one to two years, while 48.8% had implemented their program within the previous year.Regarding the source of the CPGs, 70.7% reported internal development.Of the 29.3% not internally developed or unknown, 83.3% (10 of 12) were rural hospitals or those with less than 50 beds.Of hospitals without CPGs, 96.2% (50 of 52) declared that they would be interested in assistance with implementation of a program.

DISCUSSION
A large proportion of smaller hospitals in Atlantic Canada have yet to adopt CAP CPGs.Of the different provinces, New Brunswick had the most hospitals using CAP CPGs (73.1%), while Newfoundland had the least (17.2%) (Table 2).The disparity between the proportion of hospitals identified as having a lower chance of having a CAP CPG is insufficient to explain these differences (Table 3), and likely reflects both differences between provincial styles of practice and the degree of influence of academic centres in each province.
With limitations in manpower and academic infrastructure, smaller community and rural hospitals look to academic and referral centres for guidance with reference to best clinical practice.Several authors have argued that it is indeed the responsibility of academic centres to provide such support (12,13).It would appear that with regard to CAP management, at least in Atlantic Canada, we are failing in this role.
The development of a generic CAP CPG is indicated.This needs to be flexible and adaptable to accommodate idiosyncrasies in different hospitals, areas and practice patterns where it is introduced, while still ensuring that the best clinical practice, both practical and available, is maintained.Final versions for each institution would ideally be completed in consultation and collaboration with health care providers at the target sites.Ensuring compliance with CPGs has also been found to be challenging (14)(15)(16).Evidence-based implementation strategies, including the involvement of local practitioners in the maintenance of CPG adherence, education sessions and strategies to trigger CPG use in appropriate cases, should all be offered in conjunction with the CPGs.
Weaknesses in the study include the 6.5% of cases in which the person being questioned was uncertain about whether CAP CPGs were in use at their hospital.We submit that introduction of any CPG is, almost by definition, a multidisciplinary decision, and it would be very unlikely for CPGs to be in use without the head nurse, chief of staff and hospital pharmacist all being aware of it.Presuming that 'unknown' may safely be recorded as 'no', our figures would reflect a 55.9% 'no CPG' rate.
Data from this study only demonstrate the situation in Atlantic Canada with regard to CAP CPGs, and we have no way of knowing whether this reflects the use of these guidelines in the rest of Canada.Given that the proportion of smaller hospital CPG use was similar in all four provinces surveyed, it is likely that this observation could be made in other provinces.
Because the level of CAP CPG use was lowest in the smallest (less than 20 beds) hospitals, we did not solicit information as to how many cases of CAP were treated at each hospital each year, so we cannot estimate the impact of introducing CAP CPGs to each hospital.Having said that, CPGs may be of particular use in institutions in which practitioners are not experienced in CAP treatment.
Finally, as mentioned above, simply developing CAP CPGs for use in smaller hospitals does not ensure that better processes    of care for CAP patients are followed.We did not assess implementation and maintenance strategies used, nor did we compare aspect of care or outcome variables such as length of stay, time to first antibiotic or mortality between institutions with and without CPGs in place.Such studies are indicated, as are the ongoing evaluation and refinement of implementation strategies, to ensure the use of CPGs in appropriate cases.

CONCLUSIONS
Although larger teaching hospitals are using CAP CPGs, the degree of adoption of these guidelines in smaller hospitals in Atlantic Canada is low.Efforts to produce standard CAP CPGs that can be adapted to different sites, as well as implementation strategies, are indicated.

Campbell
Can Respir J Vol 11 No 4 May/June 2004 302

TABLE 1
Reasons for exclusion of facilities