Evaluation of Health Professional Training on a Large-Scale Implementation of Chronic Care Model in Primary Care Settings of a Middle-Income Country

Corresponding Author:

Thanita Thongtan
Department of Internal Medicine
Texas Tech University Health Science Center
3601 4th Street, Stop 9410, Lubbock, Texas, USA
E-mail: thanita.thongtan@ttuhsc.edu

Abstract

Abstract

Background: The evidence on the effects of the Chronic Care Model (CCM) intervention in low- to middle- income countries is limited. A recent policy initiative in Thailand to improve care for patients with non-communicable disease focuses on strengthening primary care through the allocation of family medicine physician and multidisciplinary primary care training and allocation failed to demonstrate any changes in knowledge and skills of the trainees.

Aim: Our study aims to examine the effects of provider training and local health systems settings on provider perception of implementing CCM for patients with diabetes or hypertension in Thailand.

Methods: We conducted a cross-sectional survey of 1,064 multiprofessional providers in 128 primary care units in large and small municipalities from 11 provinces in Thailand using a self-administered questionnaire modified from the assessment of chronic illness care (ACIC) form. Responses fall within four descriptive levels D, C, B, A of implementation ranging from D "little or none" to A "fully implemented" intervention. Generalized linear models were employed to compare provider perception of implementing CCM between trained pilot PCUs, pilot PCUs, ordinary PCUs, and NCD clinics.

Results: Generalized linear models depicted an independent association between every CCM component and facility type with respondents of trained pilot PCU reporting the highest estimated marginal mean (EMM) scores (p<0.001). Statistically significant differences in the scores between health facilities in small and large municipalities were found in almost all components (p<0.01).

Conclusions: Policymakers might find the training approach promising in context with extensive universal healthcare coverage and relatively strong healthcare infrastructure like Thailand. In addition, modified ACIC in our study might be useful to assess and monitor the progress of the training in primary care settings. Further studies are needed to ascertain the effectiveness of the training using patient assessment and outcomes as indicators.

Keywords

Primary care; Chronic care model; Training effect; Physician; Hypertension

Introduction

Chronic Non-Communicable Diseases (NCD) are the major health burden globally and disproportionately affect Low-and Middle- Income Countries (LMICs) [1]. A recent systematic review found that Self-Management Support is the most frequent Chronic Care Model (CCM) intervention that is associated with statistically significant improvements in patient-level outcomes, predominately for diabetes and hypertension [2]. The CCM includes developing an organizational culture of quality and safety, support for self-management, planned interactions with the team, guidelines and provider education, follow-up care, and community linkages [3]. It is one of the most widely used models of provision of chronic care in primary care settings [4]. However, most studies on the organization of services to deliver quality chronic care are from high-income countries, and frameworks such as the CCM may not be directly applicable to LMICs [5].

In Thailand, some achievement in diabetic care at national scale was demonstrated by Kaewput et al. based on analysis of data sets from a nationwide survey conducted annually from 2010 to 2015 to evaluate the status of medical care in Type-2-Diabetic (T2DM) patients receiving care from public hospitals and clinics contracted to the largest public health insurance program [6]. The study, involving a total of 54,295 elderly T2DM patients, revealed mixed results. On one hand, it observed an increasing trend in the achievement of BP control and the use of antiplatelet medications (p for trend <0.01). Hospital admissions due to dysglycemia decreased over the study period (p for trend <0.001). On the other hand, there was a decreasing trend in the achievement of HbA1c and LDL‐C control among elderly T2DM patients (p for trend <0.001).

Apart from focusing exclusively on those conventional targets, evidence suggests achieving generic dichotomous targets (such as an HbA1c level <7%) is increasingly less relevant in contemporary practice [7]. Instead, individualized, patient-centered goals are strongly emphasized [8]. Evidence from a Randomized Controlled Trial (RCT) testing a multicomponent Quality-Improvement (QI) strategy in specialist settings in urban South Asia clearly showed improvement in patient-centered outcomes (Health-Related Quality of Life (HRQL), and treatment satisfaction) in addition to conventional outcomes (HbA1c, BP, LDLc, etc.) over 2.5 years [9]. The multi-component QI strategy comprises non-physician care coordinators and decision-support electronic health records. In agreement with earlier studies in developed country settings [10,11], the RCT study strongly supports the notions that Quality Improvement (QI) interventions directed at patients (for example, counseling), providers (for example, training), and health systems (for example, personal health record) can improve adherence, risk factor control, and patient satisfaction. Whether similar effects would be observed in primary care settings remains to be seen.

To improve care for patients with chronic diseases, a recent policy initiative in Thailand focuses on strengthening primary care with a two-pronged strategy:

• Allocation of a physician with family medicine training and multidisciplinary health professional as a team to primary care settings at subdistrict of a well-defined population of approximately 10,000

• Training of the team to deliver healthcare based on the concept of CCM. In effect, several training programs had taken place to empower primary care practitioners and leaders in Thailand to deliver care for patients with multimorbidity over the past decades [12]

However, none has demonstrated any changes in perception, knowledge, or skills of the trainees pertinent to such training. In highincome countries, several reviews of literature examined the effectiveness and implementation of healthcare team training with mixed and varying results among types of intervention and healthcare settings [13].

Our study aims to examine the effects of provider training and local health systems settings on provider perception of implementing CCM for patients with diabetes or hypertension through a cross-sectional survey.

Material and Methods

▪ Ethical approval

The present study was approved by the Institutional Ethical Review Board of the Faculty of Medicine, Ramathibodi Hospital (ID: COA. MURA2019/1018).

▪ Policy interventions

In response to a mandate of the current Constitution of Thailand (2017), the Ministry of Public Health (MOPH) adopted the patientcenteredness approach as a strategy to reform primary care. The strategy comprises

• Patient registry and engagement

• Individual care plan

• Personal health record

• Health literacy and self-management support, and

• Care continuity and coordination

Initially, 21 Primary Care Units (PCU) in 13 large (>10,000 residents) and 7 small (under 10,000 residents) municipalities in 11 provinces were chosen voluntarily to pilot test the strategy during the first year (2019) of implementation. From each PCU, the head and 2-3 clinicians attended two consecutive training workshops (1 and a half days each). The first session started with a didactic lecture addressing the concepts of the strategy and tools for translating the concepts into practices i.e., system thinking and design thinking [14,15]. The second session was two small group discussions about experiences and ideas related to the translation of the knowledge tailored to specific settings. Reading materials focusing on WHO’s Integrated People- Centered Health Service (IPCHS) and CCM were shared with the participants [16,17]. The second workshop followed one month after the first to explore the feasibility and barriers of implementing the strategy based on the participants’ try-outs. Follow up support and encouragement throughout the study period were carried out by two implementation support practitioners. They paid a visit to each team of the participants aiming at activating implementation-relevant knowledge, skills, and attitudes, and to operationalize and apply these in the context of the participants. In doing so, they aimed to trigger both relational and behavioral outcomes. For instance, the application of the concept of risk stratification of the patients was encouraged in order to customized clinical transactions according to the needs of specific patients instead of treating all patients similarly which usually results in superficial providerpatient dialogue and refilling medications over a period of just 3-5 minutes for each patient.

In parallel to the training workshop, MOPH distributed a physician trained in family medicine and new medical equipment (such as ultrasonography, ECG monitor) to each of 50 pilot PCU (those 21 PCUs included). The physician is assigned to provide full-time clinical services of 3-5 days a week to the pilot PCU in addition to outpatient care services in the referral hospital of the PCU. In contrast, patients seeking care at ordinary PCUs have only one day per week to receive care from a physician with or without training in family medicine.

▪ Population and samples

We conducted a cross-sectional survey of 1,064 providers in Primary Care Units (PCU) in large and small municipalities from 11 provinces (76 in total) in Thailand. Each of the 21 PCUs, dubbed trained pilot PCUs, was matched to a pilot PCU and an ordinary PCU in the same district. In addition, in each district, NCD clinic of the referral hospital caring for patients with diabetes or hypertension was also matched to trained pilot PCU since complicated patients from PCUs should be referred to the NCD clinic. All providers caring for patients with diabetes or hypertension in each of these facilities were asked to participate in the survey.

▪ Data collection

A self-administered questionnaire was distributed by hands to 1,064 multi-professional providers in the sampled facilities (N=128) by trained fieldworkers (nurse instructors) who described the objectives of the survey and responded to questions from the survey participants related to the questionnaire. For each PCU and NCD clinic, all the providers including the head were invited to participate by the fieldworker. It should be noted that all members of the team (including the head) were involved in patient care due to a relatively heavy workload.

The questionnaire was adapted from the Thai version of the Assessment of Chronic Illness Care (ACIC) developed by the MacColl Institute for Healthcare Innovation in the US [18]. ACIC has been tested in primary care settings of many developed countries such as the Netherlands, Switzerland, and Germany [19,20]. The Thai ACIC questionnaire was tested in primary care settings similar to those in this report with Cronbach’s alpha coefficients of 0.846 to 0.972 in each aspect of ACIC [21]. After reviewing the contents of the original ACIC and the Thai version through consultations with field experts in primary care in the study setting, we ended up with the revised questionnaire. There were only 3 sections in the questionnaire: 1) personal information of the respondents; 2) perceptions of the implementation of the strategy, and 3) perceptions of organizational support systems for the implementation. Table 1 provided details of the modified second and third sections.

Table 1: Modified ACIC questionnaire.

Responses fall within four descriptive levels D, C, B, A of implementation ranging from D “little or none” to A “fully implemented” intervention. Within each of the four levels, respondents choose one of three ratings of the degree to which that description applies. The result is a 0-11 scale, with categories within this defined as follows: 0-2 (little or no support for chronic illness care); 3-5 (basic or intermediate support for chronic illness care); 6-8 (advanced support); and 9-11 (optimal, or comprehensive, integrated care for chronic illness). Subscale scores for the six areas are derived by summing the response. Bonomi et al. presented all six ACIC subscale scores to be responsive to health care quality-improvement efforts [22]. Each team of respondents was encouraged to consult each other in the scoring, hopefully, to reach a consensus since ACIC was developed to help disease-management teams identify areas for improvement in chronic illness care and evaluate the level and nature of improvements made in their system [22-24]. Using the data from 1,064 respondents, we estimated Cronbach’s alpha of 0.92-0.93 in each aspect (p<0.05) to indicate the reliability of the questionnaire.

▪ Data analysis

All statistical analyses were performed using Statistical Packages for Social Science (SPSS) version 18.0 (IBM Corporation, New York, NY. USA).

To adjust for over or under-sampling, data were weighted to enable the representation of the target population regarding the facilities. Scores for subcomponents of Patient-Centered Care (PCC) were considered outcome variables. Professional types were regrouped into a Registered Nurse (RN), Public Health Workers (PHW), physicians, and non-physicians. To test the influence of urbanization on the outcomes, small or large municipality was applied to classify the location of the facilities where the respondents were working. Differences in the outcome variables between the facility type (trained pilot PCU, pilot PCU, ordinary PCU, NCD clinic), professional type, sex, and location of facilities were tested for statistical significance using ANOVA. Pearson’s correlation analysis was applied to test the association between the outcomes and age or duration of team membership.

The Estimated Marginal Means (EMM) and SEM of the outcomes in different predictor groups were calculated using the general linear model function. Variables were included as covariates if they met either of the following criteria: substantially modified the predictoroutcome association (≥ 10%) or significantly predicted the outcome.

Results

The majority of the respondents were female (82.9%) with a mean age of 42 years (Table 2). RN and PHW constituted over 75% of the professionals. They reported working in large municipalities at a higher proportion (55.6%) than in small municipalities (44.4%). The respondents worked in ordinary PCU at the highest proportion (32.8%).

Table 2: Characteristics of respondents at a personal level and local health systems settings (n=1,064).

Univariate analysis (weighted ANOVA) revealed a statistically significant association between facility type and health professional responses to all CCM components (Table 3). Similarly, the level of the municipality was also significantly associated with health professional responses to almost all CCM components except the involvement of specialists in improving primary care (p=0.354). In contrast, sex and professional type were significantly associated with the responses in a few components. Females gave higher scores to the implementation of informing patients about guidelines than males did (p=0.041). But males gave higher scores to the implementation of provider education for chronic illness care than females (p=0.031). Physicians gave the highest scores than other professionals in response to the implementation of information about relevant subgroups of patients needing services (p=0.046), selfmanagement support (p=0.014), and continuity of care (p=0.002). Non-physician responded with the highest scores to the question about the implementation of the regional health plans (p=0.001). Finally, the Pearson correlation coefficient revealed a statistically significant association between age and the personnel responses to the implementation of several components. Whereas the duration of PCU team membership was linked to fewer components

Table 3: Univariate analysis of the association between predictors and EMM scores (Mean (SD)) of professional responses in terms of implementation of each CCM component.

Using weighted generalized linear models with age and duration of PCU team membership as covariates in the models, we found facility type significantly affected responses to all components of CCM measures (p<0.001) in an order of highest to lowest scores as follows (Table 4); trained pilot PCU, NCD clinics, Pilot PCU and ordinary PCU. Sex, professional types, and level of municipality showed inconsistent effects on the responses across the components. Sex significantly affected responses to the implementation of informing patients about guidelines with higher scores for females. Professional types significantly affected responses to the implementation of information about relevant subgroups of patients needing services with the highest scores for RN. PHW responded with the highest score to implementation of selfmanagement support, improvement strategy for chronic illness care, and provider education for chronic illness care. Regarding the effects of the level of municipality, health professionals in smaller municipality gave a higher score for implementation of information about relevant subgroups of patients needing services, informing patients about guidelines, continuity of care, and partnerships with community organizations.

Table 4: Weighted generalized linear models for predictors and EMM scores (SEM) of professional responses to the implementation of each CCM component adjusted for age, sex, and duration of team membership.

Discussion

In this study, we reported evidence for responses of the interprofessional team of primary care providers to a policy initiative aimed at promoting CCM practices in different settings hence enabled across group comparisons. Generalized linear models depicted an independent association between every CCM component and facility type with respondents of trained pilot PCU reporting the highest EMM scores (Table 4). This indicates the possible effects of the training on providers’ reporting implementation of CCM strategy. Earlier systematic reviews of randomized trials support our findings of the possible training effects. Based on the systematic review, Dwamena F et al. concluded that interventions to promote PCC within clinical consultations are effective across studies in transferring patient-centered skills to providers [25]. Incongruent with the training described in our study, the systematic review also found short-term training (less than 10 hours) is as successful as longer training [25].

If the training effect is possible, it seems like the training affects most on the perception of freedom to address concerns of patients and families (Table 4). Given the varying degree of complexity among patients with chronic conditions, care management intensity should be tailored to reflect on the complex needs. A major challenge for this is the time constraints which is the most often cited barrier for implementing shared decision-making in clinical practice across many different cultural and organizational contexts [26]. In this regard, the team should be encouraged to use their discretionary judgment. This was, actually, a component of the training to prepare the (trained pilot PCU) team at the workshops and the follow-up visits. Apart from knowledge and skills relevant to care management, leadership support is equally important for the team to translate such knowledge and skills into actions [27]. With a realization of this leadership contribution, heads of trained pilot PCUs were included in the training.

Apart from exploring the training effects, our study also fills the gap of knowledge identified by earlier studies i.e., the influence of organizational and inter-organizational factors on CCM practices of providers. Our study demonstrated providers of trained pilot PCU reported the highest scores for every component pertaining to organization support systems: involvement of specialists in improving primary care, improvement strategy for chronic illness care, provider education for chronic illness care, and addressing concerns of patients and families. Again, these findings might be attributed to the effects of having heads of trained pilot PCUs as a trainee. Furthermore, the possible training effects on the component of organization support systems might be attributed to the inclusion of system thinking and design thinking into the training workshops. Currently, there is a paucity of evidence supporting the effectiveness of design thinking on enhancing innovation, efficiency, and effectiveness by increasing focus on the patient and provider needs [28,29].

In general, providers working in a small municipality usually face lower workload due to smaller populations as compared to those working in a large municipality. This notion helps explain our findings (Table 4) which depict reports of bigger scores from respondents in small municipality than those from respondents in a large municipality in almost every CCM component with a statistical significance: information about relevant subgroups of patients needing services, continuity of care, and partnerships with community organizations. In effect, the literature indicates that health workers are more likely to fall into traditional hierarchical practice behavior when there is a high workload.

Score differences across professional types were found to be statistically significant for many subcomponents such as regional health plans, information about relevant subgroups of patients needing services, continuity of care, etc. It is noteworthy that higher scores came from RN or PHW rather than physicians and non-physicians. To understand this manifestation, there is a need to realize that: a) under the MOPH systems the latter two groups are more likely affiliated to hospitals hosting the NCD clinics while RN and PHW to primary care units; b) the ACIC tools was designed to capture a team practice instead of individual practice. In this regard, RN and PHW with a stronger sense of a PCU team member would be more likely to respond with a higher score than physicians and non-physicians.

Conclusion

Assessing providers’ perception, our study fills the gaps of knowledge related to large-scale implementation of CCM in primary care settings of middle-income countries in terms of possible training effects on health professionals as well as varying responses to CCM components across professional type and size of the municipality. Policymakers might find the training approach promising in context with extensive universal healthcare coverage and relatively strong healthcare infrastructure. In addition, modified ACIC in our study might be useful to assess and monitor the progress of the training in primary care settings. Nonetheless, further studies are needed to ascertain the effectiveness of the training using patient assessment and outcomes as indicators.

Limitations

Our study has several limitations. First, we relied on self-report, and teams likely interpreted their performances differently. We attempted to minimize these differences through group discussion of each result. It is possible that team members did not reach consensus in determining their survey responses, allowing more assertive team members to dominate decisions. We did not take into account patient assessment of chronic care and patient outcomes. However, using modified ACIC to compare patient and provider assessment, Carryer J et al. reported a relatively high agreement between patients and providers regarding the level of self-management support received and provided. Finally, we could not compare the characteristics of the respondents with those of the target population due to the unavailability of the data. Hence representativeness of the findings might not be sufficiently addressed despite the attempt to adjust for over or under-sampling by data weighting.

Acknowledgement

This report was funded by the Health Systems Research Institute (HSRI) and National Health Security Office (NHSO).

References

1. Bennett JE, Stevens GA, Mathers CD, et al. NCD countdown 2030: Worldwide trends in non-communicable disease mortality and progress towards sustainable development goal target 3.4. Lancet 392, 1072-1088 (2018).
2. Reynolds R, Dennis S, Hasan I, et al. A systematic review of chronic disease management interventions in primary care. BMC Fam Pract 19, 11 (2018).
3. Wagner EH, Austin BT, Von Korff M. Organizing care for patients with chronic illness. Milbank Q 97, 659-664 (2019).
4. Coleman K, Austin BT, Brach C, et al. Evidence on the chronic care model in the new millennium. Health Aff 28, 75-85 (2009).
5. Beaglehole R, Epping-Jordan J, Patel V, et al. Improving the prevention and management of chronic disease in low-income and middle-income countries: A priority for primary health care. Lancet 372, 940-949 (2008).
6. Kaewput W, Thongprayoon C, Mungthin M, et al. Temporal trends in optimal diabetic care and complications of elderly type 2 diabetes patients in Thailand: A nationwide study. J Evid Based Med 12, 22-28 (2019).
7. Stone NJ, Robinson JG, Lichtenstein AH, et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: A report of the American college of cardiology/American heart association task force on practice guidelines. J Am Coll Cardiol 63, 2889-2934 (2014).
8. Inzucchi SE, Bergenstal RM, Buse JB, et al. Management of hyperglycemia in type 2 diabetes, 2015: A patient-centered approach: Update to a position statement of the American diabetes association and the European association for the study of diabetes. Diab Cares 38, 140-149 (2015).
9. Ali MK, Singh K, Kondal D, et al. Effectiveness of a multicomponent quality improvement strategy to improve achievement of diabetes care goals: A randomized, controlled trial. Ann Intern Med 165, 399-408 (2016).
10. Kerr EA, Gerzoff RB, Krein SL, et al. Diabetes care quality in the veterans affairs health care system and commercial managed care: The triad study. Ann Intern Med 141, 272-281 (2004).
11. Shojania KG, Ranji SR, McDonald KM, et al. Effects of quality improvement strategies for type 2 diabetes on glycemic control: A meta-regression analysis. JAMA 296, 427-440 (2006).
12. Suriyawongpaisal P, Aekplakorn W, Leerapan B, et al. Assessing system-based trainings for primary care teams and quality-of-life of patients with multimorbidity in Thailand: Patient and provider surveys. BMC Fam Pract 20, 85 (2019).
13. Zwarenstein M, Goldman J, Reeves S. Interprofessional collaboration: Effects of practice-based interventions on professional practice and healthcare outcomes. Cochrane database Syst Rev 8, CD000072 (2009).
14. Aboumatar HJ, Thompson D, Wu A, et al. Development and evaluation of a 3-day patient safety curriculum to advance knowledge, self-efficacy and system thinking among medical students. BMJ Qual Saf 21, 416-422 (2012).
15. Altman M, Huang TT, Breland JY. Design thinking in health care. Prev Chronic Dis 15, 180128 (2018).
16. WHO. WHO global strategy on people-centred and integrated health services. Geneva, World Health Organization (2015).
17. Bodenheimer T, Willard-Grace R. The chronic care model and the transformation of primary care. In: Lifestyle medicine. Cham Switzerland: Springer 89-96 (2016).
18. MacColl Institute for Healthcare Innovation. Assessment of chronic Illness care version 3.5. Chr Ill Care (2018).
19. Cramm JM, Strating MMH, Tsiachristas A, et al. Development and validation of a short version of the assessment of chronic illness care (ACIC) in dutch disease management programs. Health Qual Life Outcomes 9, 49 (2011).
20. Steurer-Stey C, Frei A, Schmid-Mohler G, et al. Assessment of chronic illness care with the german version of the acic in different primary care settings in Switzerland. Health Qual Life Outcomes 8, 122 (2010).
21. Gomutbutra P, Aramrat A, Sattapansri W, et al. Reliability and validity of a Thai version of Assessment of Chronic Illness Care (ACIC). J Med Assoc Thai 95, 1105-1113 (2012).
22. Bonomi AE, Wagner EH, Glasgow RE. Assessment of Chronic Illness Care (ACIC): A practical tool to measure quality improvement. Health Serv Res 37, 791-820 (2002).
23. Wagner EH, Davis C, Schaefer J. A survey of leading chronic disease management programs: are they consistent with the literature? Manag Care Q 7, 56-66 (1999)
24. Wagner EH, Austin BT, Von Korff M. Improving outcomes in chronic illness. Manag Care Q 4, 12-25 (1996).
25. Dwamena F, Holmes-Rovner M, Gaulden CM, et al. Interventions for providers to promote a patient-centred approach in clinical consultations. Cochrane database Syst Rev 12, CD003267 (2012).
26. Legare F, Ratte S, Gravel K, et al. Barriers and facilitators to implementing shared decision-making in clinical practice: Update of a systematic review of health professionals’ perceptions. Patient Educ Couns 73, 526-535 (2008).
27. Millenson ML, Muhlestein DB, O’Donnell EM, et al. Patient-centered care innovations by accountable care organizations: Lessons from leaders. Healthc 7, 30191-30194 (2019).
28. Epstein RM. Realizing engel’s biopsychosocial vision: Resilience, compassion, and quality of care. Int J Psychiatry Med 47, 275-287 (2014).
29. Carryer J, Budge RC, Francis H. Measuring care alignment in general practice consultations for people with long-term conditions: an exploratory study. J Prim Health Care 8, 256-262 (2016).