Introduction

The importance of complaints to Health Maintenance Organizations

Health organizations are called upon by many stakeholders to pay serious attention to clients’ complaints in order to uphold moral or legal obligations, to improve the quality and efficiency of health services and to maintain client satisfaction (Powers and Bendall-Lyons 2002; Schneider and Bowen 1995). Studies have shown that persons who are satisfied with their health care are less likely to sue for malpractice or to withdraw from managed care plans (Morishita et al. 1998), and more likely to seek appropriate medical care and to follow their physicians’ recommendations and prescriptions (Sherbourne et al. 1992; Ware and Hays 1988). Therefore, the reporting of complaints to regulatory agencies is commonly used as one indicator of healthcare quality (Brennan and Douglas 2002; Paterson 2002; Born and Query 2004).

From a conceptual perspective, when clients are dissatisfied or perceive a failure in service, they choose one of three courses of action (Annas 1997; Egger de Campo 2007; Hirschman 1970; Schneider and Bowen 1995; Weiser 1995). They can voice a concern by complaining to the service provider or to the relevant authorities or by sharing their negative experience with others. They can exit, that is, they cease to use the service provider. Or they can opt for silence.

Several studies show that in diverse industries, only a small portion of those who have cause to complain actually do so—choose to “voice”; those who do complain vary in how they do so and mainly lodge complaints informally, i.e., not to formal complaint handling officers, institutions, or organizations (Goodman 1999; Voorhees et al. 2006; Weiser 1995). However, while there are studies that have examined the phenomenon of complaints about health services in general (e.g., Born and Query 2004), there are not many studies that have focused on complaints made by older persons (Anderson et al. 2000; Carmel 1990; Schlesinger et al. 2002). Gal and Doron (2007) examined a national sample of persons insured by the health care system in Israel and found that 25% reported that they had had a grievance, but only one-third of this 25% had actually complained. This study was designed to go beyond the scope of the Gal and Doron (2007) research, which examined a general national sample, by employing a larger dataset that focused on complaint patterns among a specific group: older people aged 65 and above.

Israel as a comparative case-study

Israel offers an interesting natural case study in which it is possible to examine formal and informal complaints. In relative terms, Israel enjoys a high standard of health services, medical resources and research, as well as modern hospital facilities and a high ratio of physicians (3.6/1000) and specialists to the population (Ministry of Health, Israel 2004; Shalev 2003). The proportion of older people aged 65 and older in the general population has reached 10% and is experiencing more moderate aging growth (The Central Bureau of Statistics, Israel 2006). Israel enjoys a “universal coverage” health care system operated under a National Health Insurance Law enacted in 1994. The national system combines features of a single national system, such as that in the UK or Canada, with additional elements of a market-driven scheme in which multiple service providers compete, as in the USA. Within this system, Israeli citizens receive services from one of four Health Maintenance Organizations (HMOs) from which they can freely choose. Citizens also have specific legal rights to complain about their health services, e.g., at a national level to a national Ombudsman office, or to an Ombudsman of their own HMO.

Older persons and complaints to HMOs

Complaint mechanisms within the health care system are especially relevant and important to older people, who use health services more than any other segment of society (Fuchs 1999; Fried et al. 2001; Koziol et al. 2002; Bentley 2003). In Israel, the number of visits made by older persons to physicians is double that of the general population (Brodsky et al. 2007). Yet, despite the rapid growth in the older population, and the fact that this segment of the population is the main consumer of health services (and has the highest per capita health care expenditures), only one study has dealt directly with the use of complaints among older people (Anderson et al. 2000). That study found that the complaint rate of older patients in a university hospital in Australia was similar to that of younger patients. Other studies have touched upon complaints only indirectly, by looking at older persons’ satisfaction with health care providers.

In general, the findings of the different studies in this field can be divided into two categories. The first category stresses that older persons are dissatisfied with the health care services they receive and their complaints receive less attention (Carmel and Lazar 1997). Coyle (1999), for e.g., found that older people felt that the health services gave little recognition to them as human beings. These older persons also felt that they had little control over their bodies and were frustrated at being unable to gain access to care, mainly because they were not fully informed, not involved in decision making, nor were they asked questions or invited to participate in any active feedback process. Bentley (2003) also found that older people did not see themselves as consumers of health services and didn’t feel empowered to express dissatisfaction with a “free” service. They had low expectations of involvement and expressed disinterest in making their voices heard. Finally, even when older health care consumers did complain about the medical service they were in many cases ignored or not resolved by medical providers, as findings indicated that being older decreases the probability that a complaint will be resolved (Kolodinsky 1993). Dolinsky (1997) found that out of 102 older persons who did complain, 54% reported that their complaints did not result in any type of response.

The second category of studies have argued that despite high levels of unmet needs, increased age is positively correlated with patient satisfaction and negatively correlated with complaint rates (Bauld et al. 2000; Lupton 1997). However, these studies stress that other factors, which have little to do with the services received, may be influencing responses. Bauld et al. (2000) have argued that higher satisfaction or lower complaint rates can be explained by dependence on staff or programmes, fear of reprisals and lack of exit alternatives, fear of undermining the relationship with workers, fear of being perceived as “demanding”, low expectations of services, having inadequate information about the technical standards of what to expect from services (Andreasen 1985; Bauld et al. 2000; Owens and Batchelor 1996). Some scholars (for e.g., Calnan et al. 2003) have also argued that older people, because of their structural position, are characterized by passivity, acceptance, and dependency.

Walker (1999) summarized the above explanations in arguing that, in general, the biomedical model of ageing has treated ageing as a medical problem, making it synonymous with decline, and reinforcing “dependency-creating structures and attitudes”. This generalization, mirrored by an adaptive behavior of the older population, can explain the lower complaint and higher satisfaction rates of older persons within the health care system, even if there is no objective or rational justification for such behavior. This analysis is supported by the works of other scholars, for e.g., Sciegaj et al. (2004), have argued that, traditionally, older people were not empowered to direct their own care, primarily because service providers and policy makers made assumptions about their level of interest and ability to exercise control.

Aside from the potential impact of old age on the likelihood of complaining, the literature points to several key variables that may impact the tendency to complain (Stauss and Seidel 2004). In this study, we focus on two selected variables: service usage and income. As noted earlier, the literature suggests that older persons are heavier users of health services than younger persons (Bentley 2003). Hence, older persons have more interactions with health services and more opportunities to encounter service-related problems. In addition, findings from other areas imply that income also impacts on patterns of usage of health services and health status (e.g., Warren et al. 2008). Persons with lower incomes may have fewer personal resources which may reduce the tendency to complain about services.

In light of the above, the aims of this study were to examine the patterns and prevalence of complaints about health services among older clients of HMOs, explore demographic correlates, and compare results with the patterns in the younger population. In addition, the study aims to examine the associations between complaining, usage of health services, and income, an area that so far has been given little research attention.

Methods

Subjects and sampling

This study is based on the responses of subjects who participated in two national phone surveys conducted on the complaint-related behavior of clients of HMOs over a period of 3 years in Israel, from 2004 to 2007. The first survey involved a national probability sample of respondents aged 21 and over, selected randomly from a national telephone directory (see Gal and Doron 2007). Because of the relative small proportion of older participants in the general population, this study by itself could not produce a sub-sample of a size sufficient for analyzing the complaint patterns of the older population as a distinct group. Therefore, a second survey was conducted with the intent to provide further information on complaint-related variables among a larger sample of the older population in Israel. For this purpose, it employed a similar phone-based randomized sampling process, with an over-sampling of people aged 65 and above.

For methodological reasons, both surveys employed identical interviewing processes and content. For both surveys, the response rate was over 50%, with most refusals attributed to technical reasons rather than lack of interest. It should be stressed that there were no significant changes in the Israeli health care system between times that the two surveys were administered, and the statistical analysis supported the appropriateness of the pooling (for the Jewish population). We report results pertaining to Jewish respondents only, as respondents in non-Jewish communities, who comprised about 17% of the population, were not included in the second sample.

All subjects were asked, as a screening question, whether, over the past 12 months, they had wanted to complain or had a reason to complain, about any matter or problem connected with their HMO, or with a hospital where they had received treatment or another health care service paid for by their HMO. The screening was based on the subjective perception of the participants, and did not reflect an objective examination of the factual basis for the perceived complaint. This article reports results provided by 1,168 persons who responded positively to this screening question, i.e., they perceived themselves to be aggrieved by their HMO. Of these, 372 were over the age of 65 (hereinafter referred to as ‘older persons’), and 796 were in the age range 21 to 64 (hereinafter referred to as ‘younger persons’ or ‘under 65’). The four HMOs to which the screened respondents belonged were Klalit (58%), Maccabi (21%), Leumit (12%), and Me’uchedet (9%), proportions which are very close to the national HMO membership statistics for 2008 for these service providers. Analysis of the distribution of major background variables showed that, in general, the screened research sample is representative of the general Jewish population of Israel of age 21 and above.

Instruments and procedure

Most interviews lasted between 10 and 20 min. At the beginning of the phone interview, respondents who were designated based on the screening process described above as ‘aggrieved’ were then asked about the topic or problem, and if they had actually complained in any way. Those who made a complaint were further asked about how, and to which person or administrative body a complaint was submitted. Those who did not make a complaint were asked to describe their reasons for not doing so. All respondents were questioned about their knowledge of the options for complaining and their degree of satisfaction with the services of their HMO. They were also asked to provide demographic information concerning their age, gender, education, family status, and income.

The questionnaire was developed through a multi-stage pilot project and administered by the Survey Center of the University of Haifa, a professional polling unit. Interviews were conducted in Hebrew, the language spoken by the majority of the population, and in Russian, spoken by roughly 16% of the population. The questions were mostly closed-ended, and responses to all questions were recorded verbatim by the interviewers as part of a computer-aided interviewing process.

Analysis

Responses to two open-ended questions were content analyzed. Responses pertaining to perceived reasons for being aggrieved were coded based on prior work on complaining in health settings (Powers and Bendall-Lyons 2002), and classified into three broad categories: structure (for example, payments, scheduling, cleanliness, food service, administrative matters), process (for e.g., courtesy, not receiving information, communication issues and other aspects of staff–client interaction), and treatment (for e.g., misdiagnosis, errors in laboratory work, medical negligence, and other aspects of quality of medical treatment). The reasons for not complaining were thematically analysed through a multi-stage process, whereby coding categories were initially constructed jointly by two members of the research team, based on a sample of questionnaires, and then validated on further samples of responses. The actual coding of responses to all open-ended questions was done jointly by two of the researchers, who found appropriate codes and used a third member of the team when ambiguity rose.

With regard to quantitative information, χ2 tests were performed in order to find associations between complaint status (i.e., complaining or non-complaining) and age groups (i.e., under 64, 65, and above). A logistic regression was conducted to test whether complaining status (0 = complained, 1 = did not complain), could be predicted by key background variables, i.e., age, income, education, gender, and service usage, and by two interaction effects involving income and service usage. The background variables included age (0 = under 65, 1 = 65–74, 2 = 75 and above), income (0 = below national average monthly wage, 1 = average or above), education (0 = 0–11 years of schooling, 1 = full 12 years of high school and above), and gender (0 = female, 1 = male). A composite variable of service usage was measured on an ordinal scale (0 = very low, to 3 = high) to reflect the level of actual usage of health services, based on three variables: the number of visits during the last month to a physician (family practitioner or a specialist), or to hospital facilities (either outpatient clinics or hospitalization), ranging between “no visit” to “more than 3 visits,” as well as whether the respondent is taking prescribed medications on a routine basis (“yes” or “no”), as this also reflects prior interactions with health services. Data analysis was performed using SPSS version 16.

Results

Complaint patterns and associated variables

Of the 796 younger, aggrieved respondents, 33.3% (N = 265) had actually complained. In contrast, of the 372 older, aggrieved respondents, only 23.4% (N = 87) had actually complained; a difference which is statistically significant (χ 2 = 11.8, df = 1, p < 0.001). In other words, only one in every four older persons aggrieved by their HMO actually complained, compared to about one in every three of the younger persons. A further analysis showed that 25.6% (N = 56 out of 219) of the persons in ages 65–74 (‘young-old’) had complained, a somewhat higher rate (though not significantly so) compared with 20.3% (N = 31 out of 153) of those in ages 75 and above (‘old-old’). Overall, a total of 352 respondents had submitted a complaint (hereinafter called ‘complainers’).

Table 1 compares complaint rates between persons over and under-65 years of age, broken down according to several background variables. As can be seen, complaint rates were quite similar in respect to gender, education, and familial situation subgroups. However, there was a statistically significant difference in complaint rates between older persons with different income levels (below, around, and above the national average monthly wage). Complaint rates were lower when income levels were lower among older persons (χ 2 = 10.7, df = 2, p < 0.005). Older persons aged 75 years and above had a particularly low complaint rate of 20.3%, compared to a complaint rate of 25.6% among older persons in the age range 65–75.

Table 1 Complaint rates (%) for samples of respondents aged 65 and above and under 65
Full size table

Respondents were also asked to describe the reasons for being aggrieved. As explained earlier, the reasons given were classified into three broad categories: structure process, and treatment. In both older and younger groups, structural problems were the most prevalent (68 and 66%), followed by problems with treatment (25.3 and 25.6%) then with processes (21.5 and 13.0%). Note that respondents could choose more then one reason for being aggrieved, hence the percentages sum to slightly more than 100. Overall, the relative frequency of the different reasons for being aggrieved did not differ significantly between older and younger persons.

Respondents who were aggrieved, but did not complain (76% of the older persons, 66.7% of younger persons) were asked to indicate their reasons for not complaining. Responses were classified into over 10 different categories and the five most frequently cited categories are listed in Fig. 1.

Fig. 1
figure 1

Reasons for not complaining, by age (%).* Reasons marked by an asterisk are those where the difference between persons under 65 and 65 years of age and above was significant, p < 0.05

Full size image

As can be seen, the most frequent reasons for not complaining among both older and younger respondents, which together account for over 50% of all explanations, were a perceived lack of motivation or time, low expectations of getting a response from the HMO, and lack of knowledge about how to submit the complaint. There were modest, but statistically significant differences between older and younger persons in the proportions of four of the five top reasons for not complaining (marked by an asterisk in Fig. 1). Younger persons were more likely than older ones to mention the lack of motivation or time as a reason for not complaining (31.1 vs. 23.5%). In contrast, older persons were more likely than younger persons to mention poor knowledge about the existence of an option to file a complaint (19.3 vs. 12.6%), not used to complain (14.0 vs. 9.0%), and not being able to complain because of a health condition (10.9 vs. 6.6%). Both groups did not differ significantly in the proportions that noted that they did not complain due to having low expectations of getting a response from the organization (23.9% of the older vs. 26.9% of younger persons).

More than 80% of the aggrieved subjects (82.8% of older persons, 84.2% of younger persons) reported that they were not exposed to any publicity or information source about their right to complain. Only a minority (17.2% of older, 15.8% of younger persons) reported having seen sources, such as an HMO poster or leaflet, or reported receiving information from other sources, such as the media, talking to an HMO employee or a friend. No significant differences were found between older and younger subjects in the proportions exposed to information about complaints.

Method of complaining and the recipients of complaints

Complainers (N = 338) were asked how they made or submitted their complaints. One question was asked about the recipients of complaints, i.e., the people to whom a complaint was made. Replies were categorized into four types: local recipients (administrative staff, clinic nurse, etc.), senior HMO administration (such as when a client contacts regional or national management), official bodies legally designated to deal with complaints described earlier (any Ombudsman’s office or the courts), and external entities (such as consumer advocacy groups or the media). Another question asked was, “What method was used to submit the complaint?” Replies were divided into oral (face-to-face discussion, phone call) and written (letter, e-mail, printed complaint form, “contact us” form on the HMO website). The results pertaining to the above two questions are cross-tabulated in Table 2.

Table 2 Methods of complaining (represented by the recipient of the complaint), classified by age group
Full size table

With regard to the method of complaining, Table 2 shows that the majority of complaints, about two-thirds, were submitted orally for both older and younger respondent groups. Looking at the recipients to whom a complaint was made, the final column of Table 2 (labeled ‘Row %’) shows that the majority of the complaints, about two-thirds, were submitted only at the local level. A minority of the complaints were made to the HMO senior management (about 12% in both groups) or to one of the officially appointed Ombudsman bodies (14.0% of older, 21.8% of younger persons). There were no statistically significant differences between the two age groups regarding either the recipient or the method for submitting a complaint. However, when the complaint was submitted to the central administration or to an Ombudsman’s office, the proportion of written (formal) complaints was higher (p < 0.05) among the older complainers. Overall, more than half the complaints (55% of older persons and 50.4% of the younger population) were submitted only orally and at the local level.

Age as a predictor of complaining behaviour

The results presented so far suggest that older adults complain less frequently, with ‘old–old’ persons (75 and above) complaining even less than ‘young-old’ (65–75). However, it is still necessary to examine if this pattern was due solely to age or to other variables. As described in the “Introduction”, two key variables with a potential linkage with tendency to complain are income and usage of health services. In addition, other variables, i.e., gender and education, have been shown in Table 1 to also have some association with complaining status. Given the above, a logistic regression tested the effects of two interactions on complaining: age x income and age x service usage, in addition to the effects of education and gender. To control for accumulation of Type I error when there are five predictors, a Bonferroni correction was used and the criterion for significance was set at α = 0.01 instead of 0.05. (Groups used for each variable were described in the Analysis section earlier). Results appear in Table 3.

Table 3 Logistic regression to predict the probability of not complaining by age, gender, education, income, service usage, and the age × income interaction (N = 1038)
Full size table

As shown in Table 3, age, education, gender, and service usage were not found as predictors of complaint status, while income did prove to have a significant effect on complaining. The interaction between age and service usage was not significant. In contrast, the interaction between the older-old and income approached significance. Persons who are 75 years and above with below-average income had a 2.57 times higher probability of not complaining, compared to under-65 year old persons. In contrast, persons who are 65–74-years-old with below-average income were only 1.12 times more likely to not complain. Thus, it appears that the association between age and complaining shown in Table 1, i.e., that older persons complain significantly less than younger persons, needs to be interpreted with caution, as it is moderated both by the respondents’ income level and by being in the older-old age group., as well as by the conditions and correlates underlying the age–income interaction.

Discussion

Our findings are bounded by the characteristics of the national health system and culture in a single country, Israel. They are, however, based on a large, national, representative sample. Given the similarities between the health system in Israel and that in several other countries with universal health care coverage (e.g., UK, Canada, Australia or Taiwan), it is possible to make some generalizations and point to several implications for research and practice. Moreover, these points may have even stronger weight given the fact that compared to countries which do not have a universal health-care system (e.g., USA), older persons in Israel have easier access to complaint mechanisms in the system.

The finding of this study should be interpreted with caution as they are based on subjective self reporting of perceived problems and not on objective record of actual problems. The analysis treated the under 65 group as a global category, without examining possible differences between sub-groups of this segment of this population or potential cohort effects that might confound the results. Likewise, within a phone survey, which is limited by time, it is difficult to collect detailed information about health status or other aspects of the interaction with the health system. Being based on a telephone survey of a Jewish-only respondents sample, this study suffered from the natural limitation of excluding the very poor, institutionalized, and non-Jewish older populations (Brodsky et al. 2007). However, this limitation does not interfere with the conclusions, since, if these excluded populations were included in the sample, it is extremely likely that the findings would not have been weakened, but strengthened. Future research should consider how such methodological constraints can be overcome, and one possible direction may be to adopt a mixed-method approach, combining a qualitative in-depth interviews with more detailed closed-ended surveys and examination of actual records of complaints.

This study points to the existence of complaint patterns that reinforce previous findings on other populations or service contexts (Stauss and Seidel 2004; Gal and Doron 2007). In general, our findings indicate that the majority of adults who have reasons to make a complaint about health-care services do not do so. Furthermore, when adults do complain, their preferred complaint pattern was the informal and local avenue and not the formal complaint mechanism. However, the importance and uniqueness of the findings of this study are in their focus on the characteristics of the complaint patterns of older persons. The findings show that the older segment of the population, especially individuals in the low income and older-old age group, i.e., 75 and above, is a subgroup within the health care system that demonstrates a unique pattern of complaining compared to younger age groups.

The findings of this study contradict those of Anderson et al. (2000), who did not observe a statistically significant difference in complaint rates between older and younger patients in one Australian hospital. The difference in the findings can be explained by a methodological limitation mentioned by the Australian study which did not examine the reality that “…some people feel dissatisfied about aspects of hospital care but do not formally complain” (p. 411). This study, however, compared complaint rates among all participants who believed they had a cause to complain and not only those who actually complained, hence, enabling this study to expose the differences in the actual complaint patterns between the older and younger population groups. In addition this study was based on a large national representative sample of users of HMOs, as opposed to a local sample from a single hospital.

The complaint patterns observed in this study become important when one considers that compared to the general population, the usage of health care services by the older population is much higher. Although older people are more frequently in contact with the health-care system, they are less likely to complain compared to younger people. It follows that the voices of older persons are less likely to be heard or seen in any formal complaint analysis. Accessibility to complaining for older persons can be improved by more flexibility in complaint modes, such as by accepting both written as well as oral complaints (those taken over the phone or face-to-face when a complainer visits an Ombudsman’s office). These and other efforts may increase the proportion of older citizens who can exercise their legal and moral rights regarding quality health services.

The results also suggest that knowledge of rights to complain or methods in this regard is lacking in the older population, an issue that was also detected in other studies. For example, a separate study in Israel (Gross et al. 2005) found that the knowledge level about rights in the health care system was lower among older persons compared to the overall population. Likewise, Doron and Werner (2008) found that the knowledge of legal rights regarding health rights in old age were low in the older group. From a policy perspective, it is necessary to seek proactive approaches that target complaint-related knowledge of the older and poorer population within the health-care system through outreach, client education, and advocacy efforts.

Another noteworthy finding from this study is that old people are much less likely to voice concerns about issues which have subjective or intangible aspects (i.e., process and treatment outcomes), and more likely to complain about problems that can be proven or verified. Given that old persons use health services more frequently and have more medical problems than younger persons, we believe that further research is needed on specific subareas about which older people perceive reasons for aggravation, beyond the three broad categories examined here. It would also be of interest to examine the motivations behind the manner in which older people express their voice with regard to different areas of perceived aggravation, or refrain from such voicing, such as by creating friction with a health-service organization or refusing certain services or treatments (Egger de Campo 2007).

In conclusion, this study illuminates the fact that being old, especially above 75 years, combined with having a low income, hampers the ability of clients of health services to express their concerns. This in turn reduces the ability of health care providers to utilize complaint mechanisms as tools to identify needs and weaknesses in order to maintain quality of care, and as platforms to allow clients’ voices to be heard. These conclusions place new responsibilities both on the HMOs and on patients. Patients need to be aware of the need to update themselves on their rights and obligations as patients, while HMOs have a responsibility to actively engage in educational activities, such as:

  • increase the knowledge of their older clientele regarding their complaint rights;

  • make the complaint mechanisms more accessible for the older and poorer population; and

  • be proactive in their attempts to encourage and capture the voices and experiences of the older population.

From an organisational perspective, health services are increasingly called upon to pay attention to improve service quality and implement effective methods for recovery from service failures and problems (Stauss and Seidel 2004). New policies and new procedures may be needed to enable HMOs effective harvesting and analysis of the experiences and complaints of older persons, even when they do not initiate a formal complaint.