A comprehensive analysis of patients' perceptions of continuity of care and their associated factors

Abstract

Objective

To determine the patients' perceived degree of continuity of care between primary and secondary care and to identify contextual and individual factors that influence patients' perceptions of continuity of care.

Design

Cross-sectional study by means of a survey of patients attended to in primary and secondary care.

Setting

Three health-care areas of the Catalonian public health-care system.

Participants

A random sample of 1500 patients.

Main Outcome Measures

Relational, informational and managerial continuity of care measured by means of Likert scales, using the CCAENA questionnaire.

Results

Overall, 93.8 and 83.8% of patients perceived an ongoing relationship with primary and secondary care physicians, respectively (relational continuity), 71.2% perceived high levels of information transfer (informational continuity) and 90.7% perceived high levels of consistency of care (managerial continuity). Patients from health-care areas where primary and secondary care were managed by a single organization and the elderly tended to perceive higher levels of all three types of continuity. Foreign-born patients were less likely to perceive relational continuity with primary care physicians; those with higher educational levels were less likely to perceive high levels of informational continuity and patients with worse health status were less likely to report high levels of managerial and relational continuity with secondary care physicians.

Conclusions

Study results suggest high levels of perceived continuity of care, especially for relational and managerial continuity. The adopted comprehensive approach proves to be useful to properly understand the phenomenon because perceptions and associated factors vary according to the type of continuity.

Introduction

Achieving continuity of care poses one of the greatest challenges for health-care systems due to rapid advances, new treatments, high specialization and shifts in care from institutional to outpatient and home settings [1]. Therefore, patients see an ever-expanding array of different providers in a variety of places, making coordination difficult and threatening continuity of care [1, 2]. Whereas care coordination refers to the perspective of providers, continuity of care implies the result of the coordination of services from the patients' point of view and is defined as the ‘degree to which patients experience care over time as coherent and linked’ [1].

There are three interrelated types of care continuity [1]: relational continuity (RC), the patient's perception of an ongoing, therapeutic relationship with one or more providers; informational continuity (IC), the patient's perception that information on prior events is used to provide care appropriate to the patient's current circumstances; and finally, managerial continuity (MC), the patient's perception that he or she receives the different services in a coordinated, complementary and unduplicated way.

Comprehensive analyses of continuity of care are limited; the association between continuity of care and outcomes has mainly been studied with regard to RC [3] and has been associated with greater patient satisfaction, better perceived quality of life, greater use of preventive services, a higher rate of adherence to treatments and a reduction in hospitalization rates [3–6]. Furthermore, previous studies focused on a single pathology [5, 7] and generally have concentrated on analyzing individual continuity of care determinants, disregarding contextual factors [7, 8]. In spite of the fact that some studies show a strong relationship with the level of morbidity—patients with greater morbidity tend to experience lower levels of continuity of care [7, 8] —evidence regarding other individual characteristics remains inconclusive [5]. With regard to organizational factors, only one study has associated all three types of continuity with the existence of mechanisms for care coordination between organizations (health-care protocols and mechanisms to facilitate referral or sharing of information) [9].

Achieving continuity of care has also become a priority of the Spanish National Health System that is financed by taxes and decentralized into regional health services, with universal coverage and free access at point of delivery [10]. Health-care provision is organized into primary care (PC) and secondary care (SC). The main strategy employed to guarantee continuity is the assignment of citizens to a PC team that is coordinated with other care levels [10, 11]. In the Spanish region of Catalonia, the health-care system is characterized by a split of the financing and provision functions. The provision of services is the responsibility of a number of contracted providers: on the one hand, a public company, the Catalan Health Institute (ICS), and on the other hand, consortia, municipal foundations and private foundations (mainly non-profit but also for profit) [12]. This diversity, which involves a greater risk of care fragmentation, has originated different management models in the region, including the joint management of both PC and SC [13]. This makes the Catalan health system an excellent scenario in which to enhance our understanding of the phenomenon of care continuity.

This study is the first attempt to comprehensively analyze perceptions of the three types of continuity, encompassing patients from different age groups, suffering from diverse medical conditions and in different care contexts. The heterogeneity of the sample makes it possible to test the hypothesis that continuity of care varies according to the way in which health-care provision is managed, patients' socio-demographic characteristics and morbidity.

This article, which presents partial results of a wider study [14, 15], aims to determine the patients' perceived degree of continuity of care between PC and SC and to identify contextual and individual factors that influence patients' perceptions of continuity of care.

Methods

Design and settings

A cross-sectional study was carried out by means of a survey of patients in the Catalonian health-care system. The three selected areas were Baix Empordà (rural and semi-urban), the city of Girona (urban) and the Ciutat Vella district of Barcelona (urban). A single provider supplies both PC and SC services in Baix Empordà (Serveis de Salut Integrats Baix Empordà—SSIBE; an entity under private law) and in Girona (Institut Català de la Salut—ICS; an entity under public law). In Ciutat Vella, two entities supply PC (ICS and Institut de Prestacions d'Assistència Mèdica al Personal Municipal—PAMEM) and a different entity provides SC (Parc Salut Mar). The population that is at least 18 years old and is served by these organizations in the study areas are 74 144 in Baix Empordà, 83 312 in Girona and 99 093 in Ciutat Vella [16].

Study population and sample

The study population consisted of patients aged 18 or over assigned to the selected health-care areas who had received PC and SC for the same condition in the 3 months prior to the survey.

Sample size was calculated to achieve sufficient statistical power to estimate logistic regression models of Likert scales at 95% confidence level, to fulfill the de Moivre theorem of expected frequency greater than five and to express the fit and likelihood statistics as a Chi-square distribution. The sample size required was approximately 400 patients per health-care area. The final sample size was 1500, and the sample was distributed across the 3 areas according to the size of served population.

A simple random sample of patients without replacement was selected from a list of patients that fulfils inclusion criteria. This list was created from records provided by PC centers and hospitals of the health-care areas. A list of substitutes that included individuals of the same sex and age group was used to replace any refusals. Patients who had not been attended to by medical professionals or who could not understand or communicate effectively in Spanish or Catalan were excluded.

Data collection instruments and procedures

To collect the data, the CCAENA© questionnaire (Questionnaire of Continuity between Care Levels) was applied that comprehensively evaluates IC, RC and MC between care levels from the patient's perspective. This tool, previously validated in Spanish and Catalan [15], is divided into two sections. The first reconstructs the care trajectory for a specific episode, and the second, which is the object of this paper, consists of four Likert scales that measure the patients' perceptions of the three types of continuity. Two scales concern RC, the PC and SC physician–patient relationship scales that encompass the attributes of trust between providers and patients (three items), sense of clinical responsibility (one item) and effective communication (three items). The third is related to IC, the information transfer scale that includes knowledge of medical history (three items) and the supply of timely and adequate information to the patient (one item). The fourth refers to MC, the consistency of care scale that includes coordination between providers (one item) and adequate sequence of care (two items). Additionally, the reasons given by patients for their perceptions of continuity of care were collected by means of open questions. Items that constitute each scale are presented in Annex 1.

Face-to-face interviews with patients were conducted by trained interviewers, mainly at PC centers (93.7%) and also at patients' homes (6.1%) and other places chosen by patients (0.2%). Interviews were conducted in Catalan or Spanish, depending on the preference of the interviewee. Fieldwork took place between January and May 2010.

Main study variables

Outcome variables were synthetic indexes, computed from the items that constitute the Likert scales. Items had four response options that varied according to the scale: (i) strongly agree, agree, disagree and strongly disagree, in the RC scales and (ii) always, often, rarely and never, in the IC and MC scales. To estimate continuity indexes, items were scored from 0–3 (from strongly disagree/never to strongly agree/always). The simple imputation method was applied based on the mean score of the item that is considered to be adequate due to the high proportion of complete cases [17]. The second step consisted of adding the items' scores together and dividing by their highest possible score. To calculate a synthetic index of overall continuity, the previous indexes were added together, giving the same weight to each type of continuity, and then divided by their highest possible score. Finally, to simplify the analysis and the presentation of results, each continuity index was transformed into a categorical variable with four possible values (from a very high to a very low perceived levels of continuity of care) and into a dichotomous variable representing (very) high versus (very) low perceived levels of care continuity.

Independent variables were the health-care area and individual variables: socio-demographic characteristics (sex, age, immigration status and educational level) and morbidity (self-rated health status and declared number of health conditions).

Additionally, a content analysis was conducted with three open questions [reasons why patients perceive that PC and SC physicians' information was (in)sufficient and reasons for perceiving that care was (un)coordinated].

Analysis methods

The sample was weighted to adjust it to the relative size of the study population. Univariate analyses were performed to describe perceptions of continuity and reasons for it (answers to open questions). Percentages were calculated and their corresponding confidence intervals (CIs) estimated.

To assess the relationship between types of continuity and hypothesized associated factors of continuity, logistic regression models were estimated. Robust covariance adjustments—employing the health-care area variable—were used to account for correlated observations due to clustering [18]. Dichotomous variables for continuity of care perceptions were used. All the independent variables described above were included in the final analysis. Percentages and adjusted odds ratio (OR) were calculated for perceived high levels of continuity, for each scale and for the overall rating of continuity. The significance level was set at 0.05. Model fit was assessed with the Hosmer–Lemeshow goodness-of-fit test [19]. All the regression models gave P-values higher than 0.05, indicating that the model estimates fit the data at an acceptable level (Annex 2).

Statistical analyses were performed using the Statistical Package for Social Sciences version 11 and Data Analysis and Statistical Software (STATA), version 11.

Ethical considerations

The study was conducted in accordance with the current European and Spanish legislation on ethical research [20]. Informed consent was obtained from every interviewee participating in the survey, and confidentiality of data was assured by conducting the analysis anonymously. The study protocol was approved by the Ethical Committee for Clinical Research ‘Parc Salut Mar (2009/3414/I)’.

Results

More than three quarter of contacted patients (77%) refused to participate in the study. However, no differences were observed between the sample and the study population in terms of age or sex. The sample was homogeneously distributed across the different age groups; 55% of participants were female and 78% were born in Spain. About half of the patients (57%) perceived that their health status was either good or very good and 24% reported to be suffering from just one medical condition. Characteristics of the study sample are presented in Table 1.

Perceptions of continuity of care

Overall continuity was perceived as high or very high by 88.2% of patients. With regard to RC, 93.8 and 83.8% of patients perceived an ongoing relationship with PC and SC physicians, respectively. Overall, 71.2% of patients perceived high or very high levels of information transfer between care levels. Finally, 90.7% of patients perceived high or very high levels of consistency of care (Table 2).

With regard to perceptions of an ongoing relationship with physicians, information from PC and SC physicians was considered to be sufficient by 85.1 and 70.3% of patients, respectively. Reasons reported for this in open questions were that physicians provided all the information they needed (39.6 and 40.5% for PC and SC physicians, respectively), that the information provided was easy to understand (37.6 and 31.3%) and that they were able to ask all their questions (18.1 and 16.5%). Reasons for perceiving that care was insufficient were the opposite, with the exception of consultation time that was a reason reported by 10.7 and 12.2% of patients for PC and SC, respectively (Table 3).

In terms of consistency of care, 70.2% of patients perceived that care across levels was coordinated; this was due, according to open questions, to perceived communication between professionals (45.3%), satisfaction with care (17.9%) and consistency between professionals (16.9%). On the contrary, the main reasons for uncoordinated care were the perception of lack of communication (41.6%) and lack of professionalism (6.9%) (Table 3).

Factors associated with perceived continuity of care

Health-care area, age, educational level and declared number of health conditions were related to the overall perception of continuity of care (Table 4). Patients from Ciutat Vella and Girona perceived worse overall continuity of care than those from Baix Empordà [ORadj 0.35; 95% CI) (0.30, 0.40) and 0.83; 95%CI (0.75, 0.91), respectively]. Moreover, older patients were more likely to perceive a high level of overall continuity than younger patients [ORadj 3.24; 95%CI (1.51, 6.96)], and patients who had completed tertiary education were less likely to perceive a high level of overall continuity than those who had not completed primary education [ORadj 0.53; 95%CI (0.31, 0.93)]. Finally, patients with more than one health condition were less likely to perceive a high level of overall continuity than those with just one health condition [ORadj 0.74; 95%CI (0.56, 0.98)].

With regard to RC, the health-care area was associated with both PC and SC physician–patient relationships. Patients from Girona were more likely to perceive an ongoing relationship with PC physicians than those from Baix Empordà [ORadj 1.24; 95%CI (1.15, 1.35)], but less likely to give high ratings to SC physician–patient relationships [ORadj 0.38; 95%CI (0.37, 0.38)]. Patients from Ciutat Vella were less likely to give high ratings to PC and SC physician–patient relationships than patients from Baix Empordà [ORadj 0.75; 95%CI (0.70, 0.80), and ORadj 0.44; 95%CI (0.42, 0.45), respectively]. Elderly patients were more likely to report an ongoing relationship with PC and SC physicians than younger patients [ORadj 2.56; 95% CI (1.32, 4.96), and ORadj 3.66; 95%CI (3.06, 4.38), respectively]. Patients who had completed primary education were more likely to perceive an ongoing relationship with their PC physician than those who had not completed primary education [ORadj 1.51; 95%CI (1.30, 1.75)]. Foreign born patients were less likely to report an ongoing relationship with their PC physician than natives [ORadj 0.39; 95%CI (0.27, 0.58)]. Male patients, those with worse health status and those with more than one declared health condition reported less frequently to have an ongoing relationship with SC professionals than female patients, patients with better health status and those who declared to have just one health condition [ORadj 1.39; 95%CI (1.12, 1.73), ORadj 0.66; 95%CI (0.57, 0.75), and ORadj 0.71; 95%CI (0.52, 0.96), respectively].

Where IC is concerned, the perception of information transfer across care levels was associated with the healthcare area, age and educational level. Patients from Ciutat Vella were less likely to perceive an information transfer than patients from Baix Empordà [ORadj 0.24; 95%CI (0.20, 0.28)] and Girona. The elderly were more likely to report an information transfer than the younger [ORadj 4.42; 95%CI (1.87, 10.42)]. Finally, patients who had completed university education were less likely to rate information transfer highly than those who had not completed primary education [ORadj 0.57; 95%CI (0.37, 0.86)].

With regard to MC, perceptions of consistency of care were associated with the healthcare area, age and the declared number of health conditions. Patients from both Girona and Ciutat Vella were less likely to perceive consistency of care than those from Baix Empordà [ORadj 0.57; 95%CI (0.52, 0.63), and ORadj 0.63; 95%CI (0.57, 0.69), respectively]. Elderly patients were more likely to give high ratings for consistency of care than younger patients [ORadj 3.50; 95%CI (2.05, 5.99)]. Finally, patients with more than one health condition were less likely to perceive a high level of consistency of care than those with just one health condition [ORadj 0.72; 95%CI (0.56, 0.93)].

Discussion

This study represents the first attempt to analyze the three types of continuity of care across care levels as perceived by patients from different age groups, suffering from diverse medical conditions and attended to in different care settings. Analyses are comprehensive as they take into account perceived RC, IC and MC and study individual factors that may contribute to explaining continuity of care perceptions, such as socio-demographic characteristics or declared morbidity. Moreover, the selection of three health-care areas with different managerial models allows us to explore their effect on continuity of care perceptions.

Patients perceive high levels of continuity of care

With regard to RC, patients reported an ongoing relationship with both PC and SC physicians; however, they considered that they had a better relationship with the former. This result is consistent with the fact that their relationship with PC physicians is usually longer lasting [1] and patients contact them more frequently than SC, which may facilitate the establishment of stronger links [4]. On the contrary, the relationship with SC physicians tends to be more temporary [21]. Additionally, patients perceive that communication—a central element for establishing an ongoing relationship with professionals—is more effective with PC than with SC physicians in terms of completeness and intelligibility of information.

About a quarter of respondents did not perceive appropriate transfer of information, in spite of the fact that all health-care areas have implemented mechanisms to share information between care levels. Possible explanations for perceived lack of information transfer are firstly improper use by health-care professionals in terms of registering or uptaking information, and secondly that the perception of IC does not depend exclusively on results of care coordination, but on other aspects such as overall satisfaction with care. Nevertheless, this issue should be further evaluated in future research.

As far as MC is concerned, most patients perceived that care was consistent among providers. Reported reasons for patients perceiving that care was (un)coordinated are mainly associated with elements related to the organization of services such as the transfer of information and consistency of instructions among professionals. The fact that the main reason given by patients was related to information transfer implies that they perceive a strong relationship between IC and MC of care, as has been noted by other authors [22].

Health-care area and age were associated with all types of continuity of care

Patients from areas where PC and SC are managed by the same entity perceived a better continuity of care. These results seem to support previous studies which suggest that patients are more likely to perceive discontinuities, particularly regarding informational aspects, when they receive care from different health-care organizations [23]. Disparities in perceived continuity do not seem to be related to the public/private management model, but could be associated with different degrees of care coordination across levels or other characteristics of the organization of health-care provision.

With regard to individual factors, age was the only variable associated with all dimensions of continuity, with older patients reporting a high level of overall continuity more often than younger patients. Previous studies focusing on RC also conclude that older patients perceive higher levels of continuity of care [24, 25], whereas studies focusing on patients with a single pathology suggest that there is no association between MC and IC and age [5]. Moreover, previous studies have shown that older patients tend to report better experiences with care [26], may be due to the fact that they have a better knowledge of the system due to more frequent use of health-care services (for example, they are more familiar with the referral process across care levels), or because elderly patients tend to express greater satisfaction with their care received [27].

Immigration status, educational level and morbidity influence certain types of continuity of care

With respect to immigration status, foreign-born patients expressed lower levels of RC with PC physicians than natives, which is consistent with other studies analyzing the relationship between RC and ethnicity [25, 28]. However, caution is required when applying results regarding ethnic minorities to the immigrant population because the two groups of patients are not equivalent. The poorer perceptions of immigrant patients regarding RC of care could be a consequence of disparities in care due to barriers such as social prejudices of PC physicians [29]. In addition, language may act as a barrier to achieving RC [5].

Educational level showed a negative association with IC: the higher the educational level, the lower the rating for information transfer. Previous studies suggest that more educated patients judge quality of care more critically [26] and elicit more information from their physicians [30] than less educated patients. It is therefore to be expected that more educated patients identify gaps in the information transfer more easily that is in accordance with our results.

With regard to declared morbidity, results are consistent with a previous study [8] which indicates that patients with worse health status were less likely to report an ongoing relationship with physicians and less likely to report that their care was consistent. However, in contrast with other studies [7, 8], morbidity status was not associated with patients' perceptions of IC.

Limitations of the study

With respect to the limitations of this study, 77.5% of contacted patients refused to participate. Although they were replaced by others belonging to the same age group and sex, a non-response bias cannot be ruled out that would lead to the underrepresentation of certain population groups. The study population consisted of patients who had received PC and SC for the same condition in the 3 months prior to the survey and is supposedly in poorer health than the general population using health services, and therefore it is to be anticipated that the study population will perceive lower levels of continuity of care than general health-care users [7]. Another limitation is the dichotomization of indexes to facilitate the interpretation of data that could lead to some loss of information. Lastly, the scarcity of previous studies aimed at a comprehensive understanding of the phenomenon of continuity of care makes it difficult to contrast the results with those from other studies, and this makes it more likely that some relevant variables that further explain continuity of care have been overlooked in our study.

Conclusion

In summary, a comprehensive analysis of patients' perceptions of continuity of care is a useful tool for understanding the phenomenon, leading to the identification of associated factors, that have been shown to vary according to the type of continuity of care. The most consistently associated factors were health-care area and age that were related to all types of continuity of care, whereas other factors such as immigration status, educational level and morbidity were related to specific types of continuity of care. Results on identified groups of patients with poorer perception of continuity suggest the call for further research on continuity-associated factors and inform about which targeted groups would probably require the implementation of specific organizational actions to improve their care coordination. Furthermore, it should be explored if some groups, such as immigrants, are possibly affected by inequalities regarding healthcare supply. Additionally, results highlight the importance of the organization of health services with patients perceiving better continuity when they were attended to within organizational frameworks managed by the same entity, although other characteristics of health-care organizations could be mediating this relationship.

Finally, this research was conducted before two important events of the Catalan health-care system took place: the reduction in the health-care budget and the split of the major public provider into different public companies to manage separately PC and SC. As such, it will provide a base line to analyze the potential impact of these measures on health-care provision.

Funding

This work was supported by the Instituto de Salud Carlos III which partially financed the study [PI08/90154 and PI10/00348].

Acknowledgements

The authors would like to thank all interviewees for participating in the study and Miguel Martín, Silvina Berra, Maribel Pasarín, Josep María Argimon and Soledad Romea for their support in developing the study's conceptual and practical framework. Furthermore, we wish to thank Francesc Cots, Lola Bosch, Josep Maria Lisbona, Isabel Serra, Juan Antonio Camús, Mercé Abizanda, Elena Martínez and Montserrat Figuerola whose support has made this study possible.

References