Increasing importance of unmet needs for care
Our study reports high levels of unmet needs in two regions of Québec province, Canada. Nearly one person out of five reported an unmet need during the six months preceding the survey. This prevalence is higher than previously reported rates in the Canadian context for a 12 month period [1]. This difference could represent higher levels of unmet healthcare needs in the specific context under study. However, our results are in line with previous reported increases in unmet needs in the last 5 years among Canadian provinces as reported in Additional file 2: Appendix. In addition, a shorter period of recall might prove more accurate with regards to identifying instances of unmet needs. We may also have identified as unmet needs ‘problems which will eventually need to be dealt with in the following weeks’ and this might explain our slightly higher level compared to studies with longer recall periods.
Furthermore, our study highlights the importance that these unmet needs have for the public. The problems for which an unmet healthcare need is reported are perceived to have an impact on lives. This suggests that these reasons cannot be dismissed as problems that are not serious enough for the person to seek care. Attention should be paid to unmet needs by healthcare policy makers and managers as they could reflect underlying problems in access to healthcare in Canada. Recent international surveys have confirmed the acute problems of access to healthcare in the Canadian context particularly in the province of Quebec [23]. However, increasing discontent with the healthcare system and increasing media coverage about care gaps and safety problems could also encourage the reporting of unmet healthcare needs. The extent to which the increase in unmet healthcare needs results from variations in peoples’ perception, from increasing needs among the population, or from real barriers to access care remains to be assessed and further studies should try to disentangle these specific effects in understanding these phenomena. For practitioners, our study might suggest that unmet needs for care remain an important indicator of the clinics’ capacity to address their patients’ needs through the level of affiliation that they develop with their patients.
Economic vulnerability and unmet needs in a universal system
Our study is amongst the first, in the Canadian context, to identify perceived poverty as a factor associated with the reporting of unmet healthcare needs. Previous studies had suggested that the universal coverage guaranteed by the Canada health act through a taxation-based funding mechanism effectively prevents such inequity. Given the current context of increasing private costs of healthcare services in Canada, such as the private financing of diagnostic services and increasing drug costs, this result is not surprising. Access to free public healthcare services is associated with payment for drugs given on site, diagnostic procedures, or medical services in up to 28 % of cases; 33 % of people report losing revenue when consulting their primary care clinic [24]. Both utilisation and non-utilisation of healthcare are associated with costs; in our study, 20 % of people with unmet needs had lost revenues (data not shown).
Accumulated knowledge about the direct and indirect costs involved in consulting medical services could prevent poorer households to actually seek otherwise free medical services. The fact that absolute household income was rejected from the final analysis, with perceived poverty capturing most of the association, highlights the difficulty in relying on income to assess inequalities and equity. In our study, a fair proportion of people with low household income did not perceive themselves as poor. These include students and retirees for which current income is a poor predictor of economic status. A strength of this study is its use of various measures of economic status, including income and assets but also relative economic status.
Our study also found that unmet needs for care increases with educational status, in line with what was found from previous studies highlighted in the introduction. The inverse association of unmet needs with perceived poverty (the poor reporting more unmet needs) and with education (the less educated reporting less unmet needs) is also of interest. Although it is usually thought that the poor tend to be less educated, our study suggests that population studies in metropolitan areas challenge this assumption. Students and immigrants, who show higher levels of education, tend also to report lower income. The retired, even if well educated, tend to report lower income (despite accumulated assets). These complex relationships between income, education, and unmet needs could explain why perceived poverty was the most important individual economic status characteristic related to unmet needs in our study. In fact, correlations between the different economic status measures were moderate (for example, reported income moderately correlated with perceived economic status (Pearson 0.463) and reported assets (Pearson 0.339) (full data correlation matrix available upon request).
Unmet healthcare needs as a mirror of PHC organisation and accessibility
Among the strongest associations found in this study is the affiliation with a primary healthcare provider through having a family doctor. Contrasting with healthcare user studies suggesting relatively few accessibility problems in primary healthcare - at least on the basis of the geographical and temporal access measures available - our population-based study could assess the impact of not having a pre-determined entry into the healthcare system on access to healthcare at times of perceived need. This is probably where Canada, and especially the province of Québec, has been lagging behind other highly industrialised nations [25, 26]. In addition, various factors such as modifications to the age and sex composition of the medical workforce, changes in the average amount of time spent in clinical activities in primary or secondary care settings, or specific organisational characteristics of the primary healthcare infrastructure, could explain this situation.
In our study, 30 % of respondents declared not having a family doctor and up to 14 % did not identify a usual source of care in the last two years. However, the fact that people without a usual source of care also report lower unmet needs suggests that a fair proportion of people without a usual source also present few perceived care needs. This highlights the fact that, with regards to need for care, the gaps in affiliation do not necessarily translate into accessibility problems for a portion of unaffiliated people. Nevertheless, our study highlights the need, for clinical practice contexts, to foster a strong link between patients and providers in order to reduce unmet needs.
Organizational factors seem to be strong determinants of unmet needs for care as reflected by the reasons reported for not obtaining care. Three respondents out of ten identified not finding a doctor accepting new patients as a contributor to their unmet healthcare need. Other organisational factors related to unmet healthcare needs suggest that primary healthcare delivery models have a lot to do in preventing unmet healthcare needs. Affiliation with a primary care clinic seems necessary, but certainly not sufficient to ensure access. Our study did not find a difference with regards to the type of organisation reported as usual source of care, with the exception of solo providers. This might suggest that various types of organisations are able to provide an affiliation with a primary care provider, the single most important factor our study identified as protecting from unmet needs.
Our results also suggest that use of a specialist clinic or a hospital emergency department for usual care does not provide an opportunity to develop an affiliation with a primary healthcare provider and, as a result, does not protect against unmet needs for care, despite its organisational characteristics promoting constant access. Future studies could assess which organizational mechanisms could be set in place to ensure that regular users of emergency departments, or individuals for which it represents the sole source of care, are provided with alternative follow up in primary healthcare settings to prevent future unmet needs for care and consequent utilisation of emergency departments.
Our study could not find differences among specific types of primary healthcare organisations. Solo providers, private group practice, family medicine groups and local community health centres did not statistically differ from each other with regards to the proportion of patients reporting unmet needs. The relatively small number of people affiliated with local community health centres and family medicine groups probably resulted in lack of power to detect differences among these models. However, the results of the p-values also suggest that more stringent criteria to assess statistical significance would suggest no statistical difference according to usual source of care. In addition, this study could not fully assess the latter model (Family medicine groups), as it was being conducted at the start of the PHC reform. Further studies, following the more recent increase in population coverage by the emerging family medicine groups could provide stronger evidence of trends indicated in this study.
It is important to highlight here that affiliation with a primary healthcare provider is also strongly associated with age and with the types of primary healthcare organization to which people are associated. People over the age of 65 reported having a family doctor in about 90 % of cases and people reporting a solo provider as their usual source of care also have higher reporting of having a family doctor [24]. This could partly explain the fact that our study could not find differences between types of primary healthcare organizations, controlling for other factors including affiliation with a family doctor. The impact of types of primary healthcare could to a large extent be mediated through the capacity of an organisation to provide a family doctor – an affiliation with a provider that takes responsibility for your care. In fact, Family medicine groups, a model consistent with the patient-centred medical home concept, were associated with lower unmet needs in bivariate analyses. The disappearance of this association can be attributed to a great extent on the impact of having a regular family doctor. Family medicine groups and solo-providers are associated with the highest levels of affiliation with a personal primary care doctor.
Study strengths and limits
Our study is among the first population-based studies with a large representative sample to have identified and classified usual sources of care. This study provides a large (representative for sex and age composition) sample of the population enabling the reporting of rates of unmet needs and other aspects of the experience of care at the local level, and, more importantly with a link with the usual provider of primary care. Such a large sample was required to provide a good fit between surveyed organisations in the regions and the respondents selected in a random sampling fashion. Smaller sample size would not have enabled us to analyse the impact of organisational models, and related organisational characteristics, at the population level. It should also be mentioned that such a large sample size increase the risk of finding associations which in reality do not convey a clinical significance. However, the associations found in the paper are in line with previously published studies on unmet needs and the association with types of organisations, which is the distinct contribution of this paper, did not reveal significant but weak odds ratios, which would have been suggestive of a statistically significant but minor association.
The two selected regions are representative of Quebec province for age and sex distribution, are the two most populous (accounting for 43 % of the province’s population and approximately half of all primary healthcare organisations in the province) and sprawl across a diverse array of contexts, including metropolitan, suburban, small town and rural areas. The random selection of households and respondents, and the use of sampling and post-stratification weights, ensures a representative sample of the population and the provision of robust estimates whilst taking into account the sampling design. In addition, the questionnaire developed for the study is probably among the most extensive questionnaires on unmet healthcare needs currently in use.
However, our design remains a cross-sectional study. This type of study design is always subject to recall bias and difficulty in assessing the directionality of the associations found. In addition, the assessment of unmet healthcare needs remains a subjective personal evaluation, which will vary among individuals. Personal expectations towards the healthcare systems and towards health can also be associated with the perception of unmet needs, independent of people’s health status and access to healthcare services. The extent to which the reported association is the reflection of these perception biases cannot be assessed or controlled entirely, despite our use of sophisticated methods. Furthermore, other reasons for unmet healthcare that were not measured in our study may also be important and as mentioned earlier, our study occurred at the start of the reform. A longitudinal study looking at the evolution of unmet needs through several years of reform would bring a more definitive understanding of the potential impact of reform models on unmet needs.