A client-generated health outcome measure for community nursing
of life significantly more than patients with type 2 diabetes. How-
significantly to an individual’s perception of well-being and opti-
mal functioning. Measuring progress against objective clinical
criteria is appropriate for some groups of patients. However, a
large number of patients receive care for conditions where pat-
terns of illness and recovery cannot be predetermined, ongoing
care is required, and the condition impacts on areas of life in a
way that may not be directly recognised as a consequence of the
disorder. Quality of life is subjective and related to individual
circumstances, therefore the ability to determine the impact of
community health interventions using the client’s perceived quality
of life as a measure could potentially lead to appropriate and
patient-responsive interventions.
ever, the correlation of the CGI on test/re-test was more stable
with VLU. There could be reasons for this. First, the CGI scores
of diabetes patients were higher, giving more room for change
and, second, the diabetes patients had more comorbid conditions
that may have contributed to the change in the CGI in the interval
between tests. The use of Bland-Altman plots, however, shows
that the CGI is a reliable measure and that the test/re-test reliabil-
ity is not affected by the mean of the value of the CGI for either
condition.
Criterion validity was tested using an established QoL meas-
ure, the SF-36. The correlation with the dimensions of the SF-36
was found for four out of the eight dimensions for theVLU group.
This shows evidence of concurrent validity with ‘dimensions’that
are expected to be affected byVLU such as BP, PF and MH.There
is no correlation with ‘dimensions’such as vitality (VT) and gen-
eral health (GH). As such this shows evidence of discriminate
validity with constructs that are not expected to be affected by
the VLU. The low correlation in the type 2 diabetes group may be
an indication that the CGI taps into dimensions other than those
in the SF-36. This has been shown with other client-generated
instruments.7,15 In comparison the diabetes group identified more
comorbid conditions than the VLU group and this may have had
an impact on the tests. The correlation with the MH dimension of
the SF-36 by both conditions may be tapping into issues of this
age group, but this needs further investigation before conclusions
can be drawn.
Overall, this study has shown that a patient-centred measure is
a practical and feasible outcome measure to be used in a commu-
nity nursing setting. However, this cannot yet be generalised to
other conditions seen by CHNs and research is in progress to test
the utility of the tool with clients receiving care from a specialist
community service. The CGI has shown that it is stable on test/
re-test, even when clients change their nominated ‘areas of life’.
The CGI shows aspects of criterion validity and construct valid-
ity in the VLU group. The results of both reliability and validity
tests with type 2 diabetes present issues that need further investi-
gation.
The CGI also needs to be tested for sensitivity before it can be
an accepted as a measure of outcome.24 Further testing for con-
struct validity and field tests to improve its efficiency in adminis-
tration and recording are indicated. Some of these areas have been
addressed in subsequent investigations and will be reported in
due course.
To test construct validity, the CGI was correlated to selected
clinical markers. There was a significant correlation of the CGI
with pain as a clinical marker for VLU. The low numbers of re-
sponses on this parameter for the type 2 diabetes group does not
allow conclusions to be drawn.
Acknowledgements
We thank the staff in the Community Health Services in the
Illawarra, the Illawarra RetirementTrust and the participants whose
co-operation made this study possible.
Unlike an acute episode of care, which has a definite endpoint,
when the patient is discharged from the hospital, community health
services provide ongoing care to an individual, often over months
and years. Over time, the client’s need for care may change. How-
ever, in the absence of a procedure for ongoing evaluation of out-
comes and monitoring of progress, there is little evidence the CHN
can use as a basis for clinical decision making, which includes
change of interventions and discharge from the service.
The CGI provides a reliable tool that can be administered rou-
tinely to measure changes in the needs and priorities of clients.
The resulting information can be used to identify the most appro-
priate type of services that reflect the client’s priorities for im-
provement. When this type of information is aggregated at the
service level, it provides information for the design and planning
of services and service support systems.
We would like to thankAssoc. Professor Sue Kirby, Assoc. Pro-
fessor Jenny Grosvenor and Ms Sue Brown for their time on the
steering committee. Assoc. Professor Ken Russell provided the
statistical consultancy.
This research was funded by the Australian Research Council.
References
1. Australian Health Ministers’Advisory Council. Sunshine Statement. AHMAC
Health Outcomes Seminar; 1993, Canberra. Feb 3-4.
2. Marek KD. Outcome measurement in nursing. J Nurs Qual Assur 1989;
4:1-3.
3. Testa MA, Nackley JF. Methods for quality-of-life studies. Annu Rev Public
Health 1994;15:535-59.
4. Avis NE, Smith KW. Conceptual and methodological issues in selecting and
developing quality of life measures. Adv Med Sociol 1994;5: 255-80.
5. Rissell C, Ward J, Sainsbury P. An outcomes approach to population health at
the local level in NSW: practical problems and potential solutions.Aust Health
Rev 1996;19(2):23-39.
6. Lerner DJ, Levine S. Health-related quality of life: origins, gaps, and direc-
tions. Adv Med Sociol 1994;5:43-65.
7. Hurst NP, Kind P, Ruta D, Hunter M, Stubbings A. Measuring health-related
quality of life in rheumatoid arthritis: validity, responsiveness and reliability
of EUROQOL (EQ-5D) Br J Rheumatol 1997;36:551-9.
Conclusion
The notion of using an intangible and subjective construct such
as quality of life as a health outcome measure recognises
that dimensions other than recovery from illness contribute
2000 VOL. 24 NO. 5
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
533