HCA 807 Segmentation in Health Care Organizations
HCA 807 Segmentation in Health Care Organizations
People’s health and the way in which they experience it are important components or indicators of their quality of life. In current practice, a range of innovations are providing an increasing number of opportunities to tailor medical treatment to individuals. This is referred to as ‘personalized medicine’ [1, 2]. Through this approach, the effectiveness and efficiency of treatments can be increased, enabling vital improvements to these individuals’ quality of life.
In addition to biomedical care, additional supportive care programs are provided in many therapeutic areas. Such programs aim at improving therapy adherence [3, 4], intend to stimulate self-management [5,6,7], or self-care [8, 9], or may be directed at improving life-style related aspects, such as physical fitness, or diet-related recommendations [10,11,12]. There is a general agreement that programs, to be efficient and effective, should ideally suit the needs, wants and wishes of individual health-care consumers [13, 14]. The trend toward tailoring programs more closely to individuals is aligned with developments in the field of personalized medicine. ICT developments will also help to ensure that supportive programs can be personalized further in the future [15, 16]. However, it is not always clear which programs are appropriate for which individuals [17,18,19]. The allocation and the effectiveness of supportive programs can be substantially improved by the application of techniques that help to increase the fit between the programs and the intended users on a systematic, scientific foundation. Which techniques are the most suitable?
HCA 807 Segmentation in Health Care OrganizationsSegmentation
A generally accepted technique that is widely applied within marketing is consumer segmentation. Segmentation is an approach that aims at the differentiation of groups of individuals into segments [20] to align supply and demand and to facilitate the selection of the target groups [21, 22]. The principle of segmentation is directly applicable to the domain of health care in which the alignment of supply and demand is of vital concern [23, 24].
Most segmentation outcomes are empirically driven. In such instances, the segmentation criteria are based on practical considerations: a prior, post hoc, or data driven [25, 26]. Another approach is segmentation based on theoretical assumptions. This implies that the criteria for segmentation (e.g., behavior, attitudes, and beliefs) are based on/ or derived from a theoretical framework (e.g., ‘Adoption-diffusion theory’ of Rogers [27].
A major disadvantage of the empirical approach is that the use of empirical segmentation criteria may be suitable in a specific study, but problems arise when findings are generalized to other samples, other populations, or other situations or domains. This means that the segments are not stable over samples and over time. Another more fundamental problem is the fact that falsification is not possible: outcomes have no underlying theoretical framework but are exclusively based on a specific set of empirical data.
Segmentation in health care
Segmentation as a basis for allocating services, products, and information to individuals is not commonplace in health care [28]. Segmentation procedures often use therapeutic domain or stage of development of a disease as criteria for differentiation (usually exclusively bio-medically oriented). The consequence of this approach is that depending on disease type, individuals, in addition to relevant medical treatment will receive the same additional service and support. At best, there may be some differentiation depending on the stage of the disease.
Studies that include psychological factors (as a base for segmentation) are rare (e.g. [29,30,31,32,33,34,35]). It should be mentioned that all these studies are empirically driven.
A promising model for segmentation in health care is the Bloem-Stalpers model [36]. The foundations of the model consist of two extensive research projects. The first study [37] aimed at the development of a theoretical basis of the concept of subjective health. This implied a clear definition, conceptualization, and operationalization of subjective health. The second study [38] focused on the identification of the main psychological determinants of subjective health.
In his study, Bloem [37] conceptualized subjective health as an idiosyncratic and holistic concept. This conceptualization of subjective health has led to the following definition: ‘Subjective health is an individual’s experience of physical and mental functioning while living his life the way he wants to, within the constraints and limitations of individual existence’. (p. 45). (For a further elaboration on the concept of subjective health, see [36, 37]).
The study by Stalpers [38] focused on the identification of the most important psychological determinants of subjective health: (a) acceptance of the disease and/or health level, and (b) perceived control over the personal health situation. He found that the determinants acceptance and perceived control are positively correlated, and that acceptance is a stronger determinant of subjective health than perceived control.
Acceptance is the feeling of the individual that his/her health status and the possible constraints on functioning are acceptable and fitting for him/her as a person. Perceived control is the belief of the individual that his/her health status, as perceived by him/herself, can be influenced or controlled by him/herself or by others. Higher levels of acceptance and perceived control are related to higher levels of subjective health (and well-being).
The concept of subjective health, and the two determinants (acceptance and perceived control) constitute the theoretical basis of the segmentation model. As higher levels of acceptance and perceived control are directly related to higher levels of subjective health, the two determinants serve as a basis for segmentation of health-care consumers. Acceptance and perceived control serve as the two dimensions of the model, and on each dimension, two levels of scores are proposed, high and low scores. This leads to four segments of health-care consumers, with each segment representing a specific type of individual.
Based on underlying positions in terms of the two determinants, for each segment, generic psychological needs of individuals have been identified, theoretical [36, 39] and empirical [40]. It is important to note that the model is not designed for a specific patient population, but instead applies to a ‘general’ population. It is therefore based on a ‘cross-disease’ approach with a focus on individuals’ ‘health experience.
The model provides a solid framework for the segmentation of health-care consumers. First initial applications, where the framework is used, were focused on the development of a conversation approach for practice nurses [41], and the profiling of prostate cancer patients [40].
For the model to be of practical relevance, it needs further specification in terms of the following additions: (a) Further differentiation of the segments in terms the two determinants, and (b) Description and characterization of the segments in order to construct unique profiles for each segment. Consequently, the goal of the present study is twofold: the identification of criteria for differentiating between segments, and profiling of the segments of the Bloem-Stalpers model [36] with socio-economic and socio-demographic data.
In order to be able to differentiate between the segments of the model, it is essential that formal and final cut-off scores are identified on the two determinants. By determining these scores for a population of individuals (both healthy and non-healthy), a sound basis for comparison will be created for future studies. This study provides the basis for this.
To further specify the characteristics of the different segments, additional insights are needed. Over the last decades several studies have demonstrated relationships between subjective health and socio-economic and socio-demographic variables [42,43,44]. These relationships not only exist in the general population. Relations between person-related variables and subjective health have also been identified in various therapeutic areas [45, 46]. Both socio-economic and socio-demographic characteristics may be interpreted as specifications and articulations of the context within which an individual experiences subjective health, thus coloring and influencing this experience.
Therefore, a demonstration that the four segments differ in terms of socio-economic and socio-demographic variables will significantly contribute to the usefulness of the model in daily use. Advantages are that: (a) The segments can be described in more contextual detail; (b) Socio-economic and socio-demographic variables can be used to describe the context of the experience of subjective health of individuals, belonging to a certain segment; (c) These variables may serve as a source of inspiration for the development of additional support programs.
Please note that the study is focusing on the further characterization of the four segments. The study is not designed to investigate the effectiveness of supportive programs, and the study does not aim to establish the sensitivity and specificity of the classification (of four groups) to predict for example patient responses to ‘personalized medicine’ (targeted biomedical therapy).
Thus, the following research questions must be addressed: (a) How to identify formal and final cut-off scores on the determinants in order to allocate individuals to segments? (b) How is the Dutch population distributed over the four segments, given the cut-off scores? (c) How do the four segments differ in terms of socio-economic and socio-demographic variables? (d) Given the differentiation in terms of socio-economic and socio-demographic variables, which unique profiles of health-care consumers may be identified for the four segments?
Method
HCA 807 Segmentation in Health Care OrganizationsProcedure, participants and panel characteristics
In this study, a questionnaire design was used. All respondents were presented with the same questions. The data for this study was obtained by using an online survey that was given to a large panel of Dutch citizens. The panel consisted of individual members of the Dutch population who indicated that they were willing to participate in research projects. They were invited by email. Prior to acceptance, members were screened for their motivation for participating in research projects and their socio-economic, socio-demographic and residence characteristics. In addition, they were checked for duplicate panel memberships.
For all research within the research agency GfK, informed consent was treated as a formal procedure. Only respondents who declared that they had no objection to their responses being used for research were selected for the research. See GfK policy [47].
The data for this study was collected in the fall of 2012 as part of a comprehensive study into the health conditions, beliefs, values, and socio-economic and socio-demographic characteristics of individuals. During this period, the Dutch National Health system remained essentially the same in terms of internal structure and workings. Additionally, no societal volatility was observed from its participants during the past years at this point. Therefore, both the data and the outcomes of this study will describe the current circumstances in the Netherlands.
Measuring determinants
Stalpers’ questionnaire [38] for measuring the determinants was used. The questionnaire was used for several reasons. The instrument had several advantages in comparison with other general health-related quality of life questionnaires. There was a clear conceptualization and operationalization, the concept of subjective health was based on a theoretical framework, the determinants were measured with a limited set of reliable questions (only six), and in addition, the model gave direction to the kind of supportive programs individuals needed [36, 37].
The items to measure the determinants were in keeping with the world of daily experience of the respondents. Stalpers [38] stated that: ‘The data [of the qualitative study] did provide indications about the semantic characteristics of the language used by individuals … when referring to subjective health and its psychological determinants. Insights into semantics contribute significantly to the quality of the items that were used to measure the concepts, specifically to content and construct validity’. (p. 100).
Click here to ORDER an A++ paper from our Verified MASTERS and DOCTORATE WRITERS: HCA 807 Segmentation in Health Care Organizations
HCA 807 Segmentation in Health Care Organizations
Three questions were asked on acceptance of personal health condition on a scale 1 = fully agree to 7 = fully disagree, thus creating quasimetric scales. From these questions, an acceptance scale has been formed:
1.
‘I am at peace with my health condition.’
2.
‘The way in which I am functioning physically and mentally, is acceptable to me.’
3.
‘I accept my health condition the way it is.’
Three questions were asked on perceived control of health condition on a scale 1 = fully agree to 7 = fully disagree, thus creating quasimetric scales. From these questions, a perceived control scale has been formed:
4.
‘I have the feeling that I have grip on my health condition.’
5.
‘My health condition is to a great extent in my own power.’
6.
‘I have a lot of influence on my health condition.’
These six questions were presented in a random order to the respondents.
HCA 807 Segmentation in Health Care OrganizationsStatistical analyses
Factor analyses was used to summarize three questions regarding ‘acceptance of personal health condition’ to only one scale [48]. In the same way for every respondent a score corresponding to three items regarding ‘perceived control’ was calculated. Also, reliability levels (Cronbach’s alpha) of the item sets for the determinants perceived control and acceptance were calculated.
Next, a cut-off score was defined as the median value of the ‘acceptance of personal health condition’ scores. Another cut-off score was calculated as the median value of the ‘perceived control’ scores. Based on both cut-off scores four segments were constructed: Segment 1 (high score of acceptance; high score of perceived control), segment 2 (low score of acceptance; high score of perceived control), segment 3 (high score of acceptance; low score of perceived control), and segment 4 (low score of acceptance; low score of perceived control).
Finally, the socio-demographic characteristics of the patients in each segment were compared to those characteristics of the patients in other segments. By using t-tests it was analyzed whether the socio-demographic characteristics of patients in separate segments were statistically different.
Results
Sample
Upon checking the sample (a random sample, N = 2500) for the required quality demands, consisting of 2465 participants, the age distribution and the education level distribution in the sample were similar to those in the Dutch population. However, in the sample, there were more females (60.6%) compared to the Dutch population.
Acceptance and perceived control scales
The explained variance of the factor analyses model regarding ‘acceptance of personal health condition’ was 75,25%. This meant that the scale, which was calculated as the average of the scores of the three items, represented those three items very well. In the same way a scale corresponding to three items regarding ‘perceived control’ was calculated. It turned out that the explained variance of this scale was 81,71%. So, the three questions regarding ‘perceived control’ could very well be represented by this single scale. The reliability levels of the item sets for the determinants perceived control and acceptance, as expressed in Cronbach’s alpha, were high, respectively with values of .87 and .88.
Identification of cut-off points and construction of the segments
Cut-off scores were defined as the medians of the distributions of control and acceptance, respectively. As approximately one third of the sample consisted of males, as compared to an expected distribution of 50% of males in the population, an additional procedure had been carried out to check whether this characteristic had an effect on the values of these medians. To that end, the medians for control and acceptance in the population were calculated through the use of interpolation of the cumulative percentages for males and females, based on the starting point of an equal share of males and females. This procedure demonstrated that the cut-off points of the population-based medians were similar to the sample-based medians. Based on this outcome, the sample-based medians were accepted as true and reliable cut-off points.
The median value of perceived control is 5.36; the median value of acceptance is 4.96 (both on 7-point scales).
Based on the cut-off points four segments were then identified as follows:
Segment 1: average score of acceptance > 4.96 and average score of perceived control > 5.36;
Segment 2: average score of acceptance ≤4.96 and average score of perceived control > 5.36;
Segment 3: average score of acceptance > 4.96 and average score of perceived control ≤5.36;
Segment 4: average score of acceptance ≤4.96 and average score of perceived control ≤5.36.
The sizes of the four segments were: segment 1: 31.8%, segment 2: 17.4%, segment 3: 18.9%, and segment 4: 31.9%. The largest were the segments in which the levels of control and acceptance were both high (segment 1) or both low (segment 4). This indicated a positive correlation between perceived control and acceptance.
HCA 807 Segmentation in Health Care OrganizationsSocio-economic and sociodemographic description of the segments
First, the scores on the categories of the socio-economic and socio-demographic variables were coded by the numbers 1, 2, 3, etc. (Table 2). Next, the average scores of the variables per segment were calculated and described in Table 2. By using t-tests was tested, whether the average scores of the variables were different (on a 5% level).
References
1.
Sychev DA, Malova EU. Evidence-based pharmacogenetics: is it possible? Int J Risk Saf Med. 2015;27(1):S97–8 Available from: https://doi.org/10.3233/JRS-150706.
PubMed
Article
Google Scholar
2.
Cesuroglu T, Syurina E, Feron F, Krumeich A. Other side of the coin for personalised medicine and healthcare: content analysis of ‘personalised’ practices in the literature. BMJ Open. 2016;6(7):e010243 Available from: https://doi.org/10.1136/bmjopen-2015-010243.
PubMed
PubMed Central
Article
Google Scholar
3.
Ganguli A, Clewell J, Shillington AC. The impact of patient support programs on adherence, clinical, humanistic, and economic patient outcomes: a targeted systematic review. Patient Prefer Adherence. 2016;10:711–25 Available from: https://doi.org/10.2147/PPA.S101175.
PubMed
PubMed Central
Google Scholar
4.
Agarwal P, Poeran J, Meyer J, Rogers L, Reich D, Mazumdar M. Bedside medication delivery programs: suggestions for systematic evaluation and reporting. Int J Quality Health Care. 2019;31(8):53–9 Available from: https://doi.org/10.1093/intqhc/mzz014.
Google Scholar
5.
Van Vugt M, de Wit M, Cleijne W, Snoek FJ. Use of behavioral change techniques in web-based self-management programs for type 2 diabetes patients: systematic review. J Med Internet Res. 2013;15(12):e279 Available from: https://doi.org/10.2196/jmir.2800.
PubMed
PubMed Central
Article
Google Scholar
6.
Whitehead L, Seaton P. The effectiveness of self-management mobile phone and tablet apps in long-term condition management: A systematic review. J Med Internet Res. 2016;18(5):e97 Available from: https://doi.org/10.2196/jmir.4883.
