Abstract
Consumer contributions in influencing the establishment and provision of health services have been increasingly recognised as important over the past 30 years. Much of the work exploring consumers' views on their health care has focused on their overall satisfaction with the services they obtained. However, while this global measure allows some assessment of overall health services, it gives minimal information to policy makers, service providers and administrators on how such services can be improved [1, 2]. Consumers who report satisfaction with mental health services do not necessarily report receiving enough of particular types of help [1, 3, 4]. Examination of consumers' reports of met and unmet need using a framework that covers different aspects of the mental health service, for example, the provision of information, medication, psychological therapy, social interventions and skills training, will provide policy makers with better information on how and where such types of help should be furtherdeveloped [5]. Numerous factors may contribute to whether consumers report met or unmet need for types of mental health help. In exploring factors related to consumers' reporting unmet needs, we have drawn on the Andersen and Newman [6] framework, which distinguishes three classes of factors: need, enabling and predisposing factors. This framework has previously proved useful for identifying predictor variables related to the use of health services for mental health problems [7]. We have extended this framework to include additional factors that may relate to consumers' reports of unmet needs. This broader group of factors can be considered in two categories relating to the consumers of the services and to the health system.
Consumer-related factors
Consumer related variables can be considered in three categories: (i) measures of psychological distress and mental disorder that affect the consumer's need for such services; (ii) factors relating to the consumer's ability to obtain appropriate services to address those needs; and (iii) factors that may affect the consumer's self-perceived needs for such services or their perspectives concerning whether those needs have been met.
Need factors
Need factors would include clinical diagnosis of mental disorder or the consumer's self-identifying as having a mental disorder. It might be assumed that the greater or more numerous an individual's mental health care needs, the more likely that some of those needs will be reported as unmet.
Enabling factors
This second group of factors tallies with the enabling factors identified by Andersen and Newman [6]. This group would include factors affecting the individual's awareness and knowledge of mental health problems and possible treatments, and also factors affecting their ability to present this problem in the clinical setting: for example, education, previous use of mental health services or experience of others within their social network using such services [8–10]. It would also cover factors that may affect the accessibility and affordability of such care: level of education, whether usual language was English, geographical location and income measures, such as whether unemployed.
Self-perception of need factors
There is a third group of factors that should also be considered when examining consumers' reports on whether or not their health care needs have been met. In answering questions on whether they received sufficient health services, consumers will be making their own subjective assessment of whether the help provided met their self-perceived mental health needs. Various factors that may not be related to the adequacy or efficacy of such care may nonetheless contribute to a consumer's reporting unmet need for mental health services. Some of these factors would align with the predisposing factors identified by Andersen and Newman [6] as predictors of higher levels of utilisation. These would include sociodemographic variables such as age and sex that would not, by themselves, be expected to affect need for mental health help. Individuals with particular health or personal problems may also seek types of help not usually offered by the practitioner, or considered by the practitioner to be inappropriate; for example, those seeking complementary therapies for a mental disorder. There is also a dispositional factor that may affect whether the individual reports mental health care needs as being met. Respondents who bring a ‘plaintive set’ to their life experiences may also report that their mental health needs were not met [11].
Health practitioner factors
A range of variables relating to the structure of the health system and health care providers may affect the extent to which individuals report their mental health care needs as being adequately met. Health practitioners from whom help is sought need to be able to recognise mental health problems, to understand the presenting problem of a particular patient and to offer appropriate types of help. Practitioners who have limited expertise in mental health issues, inexperience in dealing with distressed patients, who are unable to provide particular types of care (for example, not legally able to prescribe medications), or unwilling to provide care that is seen as time-consuming and financially unattractive, will all reduce the likelihood that their patients' mental health needs will be met.
Unmet need for mental health help in Australia
Previously there has been limited scope for exploring unmet need for mental health services in Australia. However, this deficit has been addressed by information collected through the National Survey of Mental Health and Wellbeing [12]on utilization of health services for mental health problems and on consumer-identified partially met or unmet need for such services. This survey of a representative sample of Australians aged 18 and over living in the community was conducted in 1997 by the Australian Bureau of Statistics (ABS). The range of information collected covered sociodemographic, physical and psychological health, levels of utilization of general and specialist mental health services, and types of help provided as part of that treatment. Five types of help were considered: information about mental illness and its treatment, medication, psychological therapy, social interventions (help to sort out problems concerning housing or money) and skills training (help looking after oneself or one's home or improving one's ability to work). For each of these five types of help, respondents who reported receiving such help were asked whether they considered that they had received enough, while those reporting not obtaining such help were asked if they felt that they had needed it.
Meadows and co-authors have previously described the prevalence of perceived mental health needs from these survey data [13]. They estimated the extent to which such needs were reported as fully met, partially met (that is, received some, but not enough, of that type of help) or unmet (that is, having not received any of a particular type of help, although it was considered to be needed). Meadows et al. estimated that 13.8% of the Australian population had a perceived need for mental health care of any type, with 41.3% of this subgroup reporting all of their needs being met. For those meeting criteria for diagnosis of mental disorder, 94.5% identified themselves as needing mental health helpwith 47.3% of this subgroup reporting that their needs were fully met. Similar analyses of perceived and met need for each of the five types of mental health help were also undertaken and reported [13].
What is covered by ‘unmet need’?
Those reporting unmet need for mental health help can be considered as two subgroups: those who have not yet obtained any help from the formal health care system and those who have obtained some mental health help but still consider that their needs remain unmet or only partially met. The analyses reported in this paper focus on this second group: consumers who have already entered the formal health-care system, but report having unmet or partially met needs for different types of mental health help.
Through analyses of the National Survey, we sought to identify the extent to which consumer- and health practitioner-related variables were associated with consumers of health services for mental health problems reporting that their self-assessed needs were met, were only partially met, or were not met at all. Four categories of predictor variables that might be associated with reporting unmet need for mental health help were explored. Three of these related to consumer variables: need for such services, enabling factors and self-perception factors. The fourth category of health system predictor variables related to the type of health practitioner seen.
In order to examine these associations, we undertook exploratory analyses using simultaneous multiple ordered logistic regressions. This method of analysis allowed us to assess the extent to which need, enabling, self-perception of need and health practitioner factors were associated with reporting unmet need while controlling at the same time for other predictor variables.
Method
The National Survey
The National Survey was conducted throughout Australia from May to August 1997 on a voluntary basis. The sample selected for the survey involved residents of private dwellings. Those in special accommodation or dwellings such as hospitals, institutions, nursing homes, hostels and hotels were not included, nor were homeless persons or those from overseas holidaying in Australia, members of non-Australian defence forces and their dependants, and households containing non-Australian diplomatic personnel [12]. Approximately 13 600 dwellings were approached with one person aged 18 years or over in each dwelling randomly chosen to participate in the survey. A total of 10 641 persons completed the survey interview giving a response rate of 78%.
Measures obtained
Sociodemographic details collected from each participant included items covering age, sex, marital status, household structure, languages used, level of education attained and labour force details. Detailed information was collected on the mental health of each individual using the Composite International Diagnostic Interview (CIDI), a computerized version of which (the CIDI-A), including diagnostic algorithms, was developed for this survey by the World Health Organization Training and Reference Centre in Australia. While the reliability and validity of the CIDI have been assessed by a number of studies [14] other writers have expressed concern about the potential for this instrument to provide overestimates of the prevalence of mental disorder in the community [15].
Following completion of the CIDI-A, survey participants were asked to identify the self-reported health problem that they considered troubled them the most. This allowed identification of those participants who self-reported a mental health problem as their main health problem.
In this analysis, participants were classified as having a CIDIdiagnosed affective disorder, anxiety disorder or substance-abuse disorder if they were given one or more of the relevant ICD-10 codes. These broad categories of mental disorder were selected over more specific diagnoses since many of the latter have low prevalence rates and high comorbidity with other diagnoses in the same category, thereby limiting their usefulness in analyses of community-based surveys. Self-identified mental disorders were similarly classified into categories of depression, anxiety and substance abuse.
Respondents were then asked about their levels of health service utilization, the types of health practitioners from whom they had obtained mental health care in the previous 12 months and the types of help they had received for any mental health problems during that period. For each type of help (information, medication, psychological therapy, social interventions and skills training), those receiving such help were asked if they had received enough, while the remainder of those who had obtained any mental health help in the previous year but not this type of help, were asked whether they thought they had needed it.
Statistical methods
Ordered multiple logistic regressions were used in these analyses with each dependent variable taking three possible values relating to the extent to which consumers reported having unmet or partially met needs. For each of the five types of help, those who received enough of that type of help were scored zero, those receiving some but not enough were scored one, and those who received none of that type of help, but thought they had needed it were scored two. However, the numerical values of 0, 1 and 2 have no significance apart from determining the ordering of the three outcomes. Predictor variables included enabling and self-perception of need measures, types of health practitioner seen, CIDI-diagnosed mental disorders and self-identified mental health problems (see Table 1). Health practitioners were considered in four separate categories: general practitioners, psychiatrists, psychologists and other health professionals, with the last category covering drug and alcohol counsellors, other counsellors, nurses, mental health teams, chemists and ambulance officers.
Descriptions of predictor variables
By applying simultaneous multiple ordered logistic regressions in our analyses we could identify any need, enabling, self-perception of need, and health practitioner factors that were associated with consumers reporting unmet or partially met need for mental health help.
Finally, each participant's survey information included weighting factors provided by the ABS. These factors gave survey estimates conforming to independent estimation of the Australian population during the time of the survey and allowed reliable, population-level estimates of variables to be calculated. Analysis was undertaken using the statistical package STATA Release 6.0 [16].
Results
The number of respondents who reported obtaining any mental health help was 1329. This subgroup of consumers represented 12.5% of the weighted population. As shown in Table 2, of all five types of mental health help, need for psychological therapy was most frequently reported as not being fully met. Of those who received some psychological therapy, 11.4% reported that their needs were only partially met. In addition, 12.5% of consumers who received some mental health help but not psychological therapy considered that they had needed this type of assistance. Conversely, 92% of consumers reported their needs for medication as being fully met. We then undertook simultaneous multiple ordered logistic regressions to identify predictor variables related to reporting partially met or unmet need for each of the five types of mental health help. These results are given in Table 3 and 4.
Percentages of those receiving any mental health help who reported that their need for each type of help was met, partially met or unmet
Odds ratios for associations of reporting unmet need for mental health help with measures of need, self-perception of need, enabling and health practitioner factors
Odds ratios for associations of reporting unmet need for mental health help with measures of need, self-perception of need, enabling and health practitioner factors
Unmet need for information
As seen in Table 3, simultaneous multiple logistic regression identified no enabling or self-perception of need variables, but one health practitioner variable, to be associated with this unmet need: having seen a general practitioner. Of the measures of mental disorder and self-identified mental health problems included in the analysis, we found those who self-identified as having anxiety were also more likely to report that their need for this type of help was not fully met.
Unmet need for medication
Only one enabling variable was found to be associated with reporting unmet need for medication: not having or undertaking higher education. No self-perception of need or health system factors were so associated. Those meeting criteria for a CIDI diagnosis of anxiety disorder also reported unmet need for this type of help.
Unmet need for psychological therapy
As seen in Table 3, multiple ordered logistic regressions indicated that only one enabling variable was significantly associated with unmet need for psychological therapy: not living in a rural location. Two need-related factors were found to be associated with reporting unmet need for psychological therapy: self-identifying as having depression and self-identifying as having anxiety.
Obtaining social interventions
Four predictor variables covering enabling, self-perception of need and health practitioner factors were found to be significantly associated with reporting unmet need for social interventions: being divorced, receiving a government pension, seeing a general practitioner or other health practitioner. No measures of need were found to be so associated. These results are given in Table 4.
Unmet need for skills training
Two predictor variables measuring self-perception of need were found to be associated with reporting unmet need for skills training in the multiple logistic regression: being male and living alone. No enabling or health practitioner factors were significantly associated with this measure of unmet need. Two measures of mental disorder were also associated with reporting unmet need for skills interventions: having CIDI diagnosis of affective disorder or self-identifying as having anxiety.
Discussion
In this analysis we examined associations between variables measuring enabling, self-perception of need, health practitioner and need factors, and consumers reporting that their needs for mental health help were unmet or only partially met. We considered five different types of mental health help: obtaining information or medication; receiving psychological therapy; obtaining social interventions relating to house and money problems; and receiving skills training to assist in looking after oneself or one's home, or in improving one's ability to work or use time in other ways. Our dependent variables measuring unmet need for each of the five types of help took three possible scores. These scores related to reporting having needs met for that type of help; obtaining some but not enough of that help; and receiving no such help but considering that it was needed. Using these three-value dependent variables, we carried out multiple ordered logistic regressions. These analyses allowed us to explore whether predictor variables associated with reporting unmet need for mental health help were confined to those measuring the need for such help, or whether such associations were also found for variables measuring enabling, self-perception of need and health practitioner factors.
Enabling factors associated with unmet need
Enabling factors associated with unmet need could affect the individual's ability to recognize problems, their knowledge of the appropriate forms of help for such problems as well as difficulties in accessing such help. Our results suggest that those with lower levels of education had expectations of receiving medication for their mental health problems that were not addressed. It cannot be determined from this analysis whether such respondents perceived that they needed more medication to treat mental health problems that cannot be effectively treated with medication or to reduce other health problems (for example, problems sleeping, relief from somatic symptoms).
We also found unmet need for social interventions, concerning housing and money problems for those whose main income was some form of government pension. While such benefits usually entitle recipients to subsidized medical care, other health and living expenses may present particular financial hardship for this group.
Finally, we found living in a rural location to be significantly negatively associated with reporting unmet need for psychological therapy. This finding would also indicate that unmet or partially met need for such help is more likely to be reported by those living in more populated areas. Further exploration of this issue is required in order to identify specific locational factors that are associated with reporting this unmet need.
Factors affecting self-perceptions of need associated with reporting unmet need
There were surprisingly few significant findings to indicate that factors affecting self-perception of need were associated with reporting unmet need. Men were more likely to report an unmet need for skills training; that is, caring for self, home or improving one's ability to work. We also found those living alone to report unmet need for this type of help. Both men and women living alone could well need help caring for themselves or their homes, and it could also be expected that men in these circumstances would have a greater need for help, given their customary role in our society. Given the ongoing reconstruction of the Australian manufacturing and farming sector [17], it could be expected that men who have become unemployed and potentially unemployable, would have a greater need for help to improve their ability to work. While there are likely to be other community resources available to help deal with both types of problems, these are unlikely to be directly available from a health practitioner.
We also found that divorced respondents were more likely to report unmet need for social interventions. This finding is not unexpected, given that couples who are divorcing often face particular financial difficulties given that they are likely to require additional accommodation and transport. Again, these are services that are more likely to be available through the non-medical, community government services.
Type of practitioner seen associated with unmet need
Those seeing general practitioners and other health practitioners reported receiving fewer social interventions than they felt they needed. This type of help covered advice on housing and money matters and is not especially the responsibility of general practitioners or health practitioners in general. This finding may indicate that people have difficulty obtaining such advice from other community based, non-medical community information and referral services, possibly because they are less aware of these non-medical community resources, and unrealistically see the general practitioner or other health adviser as an overarching source of advice on general problems.
Those seeing general practitioners for mental health reasons were also more likely to report unmet need for information on mental health problems and their treatment, although visiting such practitioners has been previously found to be positively related to receiving this type of help [18]. This would suggest that general practitioners and those using their services have quite different perspectives on the appropriate forms of help to be delivered. Our analyses, however, did not explore unmet need associated with obtaining different combinations of service utilization (e.g. seeing a general practitioner and a psychiatrist). Hence, we cannot identify here the level of unmet need that could be attributed specifically to having visited a general practitioner.
Measures of need associated with reporting unmet need for help
There is not necessarily a simple relationship between having a clinically diagnosed or self-identified mental health problem and considering that one has obtained enough help. Those with more severe conditions and hence greater need for help may be more likely to reach a threshold severity level which allows them more direct access to a range of services not offered to those with subthreshold, but still treatable, symptoms [19]. The type of mental health problem occurring (for example, whether anxiety or substance-abuse disorders) may also affect whether the individual seeks or avoids particular types of help and also whether efficacious help of that type is available for their condition. These factors may have contributed to our findings that those with selfidentified anxiety report unmet need for information about their condition, for psychological therapy and for skills training, while those with clinical anxiety considered they needed more medication. More detailed exploration of these findings is needed to identify the specific issues and problems experienced by those with clinical and self-identified anxiety when they seek mental health help.
We found self-identifying as having depression to be associated with reporting unmet need for psychological therapy. This group of respondents may well find that counselling and other forms of psychological therapy, being more time-consuming, are less easily available from the general practitioner, more difficult to arrange and, if obtained from a psychologist, expensive as well.
Finally, we found having a clinical diagnosis of affective disorder to be associated with reporting unmet need for skills training. Possible reasons for this particular finding are not immediately evident and further analysis would need to be undertaken if this result is to be explained.
Limitations of the study
Our study has considered unmet and partially met need reported by those who had already obtained some form of help from the formal health care system. In this paper, we have not explored predictor variables associated with reporting unmet need by those who have not yet received any health services.
As previously noted, the analyses reported in this paper are largely exploratory. We sought to identify enabling, self-perception of need, health practitioner and need factors that might be associated with Australians reporting unmet need for mental health help. However, our analyses did not consider more specific hypotheses that could identify individual predictor variables that might be so associated. As such, our findings should be seen as preliminary to further work on this important topic of unmet need for mental health care.
Finally, we recognize that some of our findings may be the consequence of our exploring this large sample, which may have resulted in relatively small effects achieving statistical significance.
Conclusion
In this study we explored predictor variables measuring enabling, self-perception of need, health practitioner and need factors that were associated with consumers reporting unmet or partially met need for particular types of mental health help. An important finding from these analyses is that those who self-identified, or were clinically diagnosed, as having anxiety reported partially met or unmet need for four of the five types of help: information, medication, psychological therapy and skills training. Some of our findings also raise concerns about potential lack of collaboration between health practitioners and other community resources which might be able to provide social interventions and skills training to those needing such help. This lack of collaboration between service sectors has been recognized previously as a problem for those working in general practice [20].
Overall, our results highlight the need for further analyses to be undertaken if we are to have a clearer understanding of why particular subgroups of Australians, who have already accessed formal health services, continue to report unmet or only partially met need for mental health help.
Footnotes
Acknowledgements
The authors would like to thank Keith Dear, Ailsa Korten and Scott Henderson for their help in preparing this paper.