Introduction

Immunization is one of the most cost-effective public health interventions. Yet, vaccine uptake remains below target in many low- and middle-income countries for reasons beyond vaccine supply or access. Communication between health workers and communities also shapes whether parents accept, delay, or refuse vaccination (1). Indonesia’s national Expanded Program on Immunization targets 95% full basic immunization coverage, but coverage varies widely between districts (2). In 2024, Tana Toraja District in South Sulawesi reported full basic immunization coverage of only 43.06%, the lowest in the province (3), suggesting that efforts to improve coverage should address not only logistics but also the quality of immunization education provided by frontline health workers.

Interpersonal communication (IPC) is central to immunization education. Health workers are often parents’ first source of information, especially when families have concerns about vaccine safety, previous adverse experiences, or conflicting information from their social environment. IPC interventions can improve immunization coverage and reduce vaccine hesitancy in resource-constrained settings (4, 5). However, many primary care workers have limited access to structured communication training. Consequently, immunization counselling often remains information-heavy and one-directional, with limited opportunity to listen to parents’ concerns or explore the cultural and emotional factors underlying vaccination decisions (6, 7).

Sociocultural context is important when designing communication training. Relationships between health professionals and patients in Indonesia are often hierarchical, while family and community values strongly influence health decisions. Communication models developed for more egalitarian settings may not, therefore, fully address these interactional patterns (8). The Greet–Invite–Discuss (GID) Guideline, developed by Claramita et al. at Universitas Gadjah Mada, supports two-way, empathic, and partnership-oriented communication in hierarchical and community-oriented Southeast Asian healthcare contexts (9). Greet establishes an empathic and trust-based opening; Invite elicits caregivers’ perspectives, concerns, and beliefs; and Discuss facilitates two-way explanation, shared decision-making, and follow-up planning (9, 10). Previous GID studies were conducted primarily in Yogyakarta and Central Java (8, 9). Therefore, its direct transferability to Tana Toraja should not be assumed. This study evaluates the initial application of GID in a distinct Indonesian sociocultural setting, although formal cross-cultural validation was beyond its scope.

Training programs also require an appropriate evaluation framework. The Kirkpatrick model evaluates training outcomes across four levels: Reaction, Learning, Behaviour, and Results (11). For a single training event, Levels 1–2 provide a feasible starting point for assessing participants’ immediate reactions and learning before progressing to workplace behaviour or program-level impact (12). To our knowledge, this is the first published evaluation of GID-based communication training for immunization officers and the first reported integration of Gagné’s Nine Events of Instruction into a GID-based training design. The present training operationalized Gagné’s framework by progressing from an attention-gaining video, learning objectives, and activation of prior experience to concept presentation, stage-specific demonstrations, guided role-play, feedback, reflection, and application planning (13). The Methods section maps the complete instructional sequence event by event (Table 1).

This study evaluated a one-day GID-based IPC training program for immunization officers in Tana Toraja District, Indonesia. Specifically, the objectives were to 1) describe the participants’ reactions to the training (Kirkpatrick Level 1); 2) determine pre-to-post changes in overall and domain-level GID knowledge (Kirkpatrick Level 2a); and 3) describe immediate post-training communication skills achievement across the three GID domains (Kirkpatrick Level 2b). An additional exploratory analysis examined the association between knowledge change scores and post-training skills achievement.

Methods

Study Design and Participants

This primarily quantitative, single-group program evaluation was conducted in Tana Toraja District, South Sulawesi, Indonesia. Kirkpatrick Levels 1 and 2 assessed immediate reactions, pre-to-post knowledge change, and immediate post-training communication skills achievement. Because the study lacked a comparison group, pre-training skills assessment, and follow-up, the evaluation was not designed to establish causal effects, skills improvement, retention, or workplace behavioural change (14).

Sampling was a facility-level census; all 23 Puskesmas and one private maternal and child hospital in the district program each designated one immunization representative. Participants were not randomly selected within facilities. Thus, 24 facility-designated officers were enrolled, ensuring representation of every facility but not necessarily reflecting within-facility variation or overall communication capacity.

Eligibility required active involvement in immunization services, a formal institutional assignment letter, willingness to complete all procedures, and written informed consent. We excluded officers who participated in formal communication training or communication-focused continuing education within the previous six months. The participants who did not complete a primary evaluation component were excluded from analysis; all 24 completed the evaluation and were analysed.

Training Program

The one-day, face-to-face program was conducted at the Tana Toraja District Health Office on 11 April 2026. Registration, informed consent, the knowledge pre-test, and the formal opening took place from 08:00 to 09:00 local time (UTC+8). The 09:00–14:30 core block provided 265 contact minutes: 110 minutes of didactic instruction and facilitator demonstration, 90 minutes of structured role-play, and 65 minutes of orientation, reflection, feedback, and application planning. The didactic/demonstration-to-role-play ratio was 55:45. The block also included a 20-minute coffee break, a 30-minute lunch and prayer break, and approximately 15 minutes for transitions and post-training assessment preparation.

The program was an introductory district-level workshop, not a course intended to establish mastery in one session. The principal investigator, a physician with a master’s degree in Medical and Health Professions Education, facilitated the training under the academic supervision of two senior medical-education faculty members. Before implementation, medical-education and clinical-communication experts reviewed the slides, knowledge test, reaction questionnaire, role-play scenario, and GID checklist.

Materials comprised an attention-gaining video, expert-reviewed slides, a standardized printed role-play scenario, and a behaviourally anchored GID checklist. The curriculum covered IPC principles, the GID Guideline, cultural sensitivity in immunization education, vaccine hesitancy, and shared decision-making. Gagné’s Nine Events were operationalized through attention gaining, clarification of objectives, activation of prior experience, concept presentation, stage-specific demonstration, guided role-play, feedback, reflection, and application planning (Table 1).

The standardized scenario, “MR Immunization Education for a Hesitant Mother,” presented a caregiver concerned about measles–rubella vaccination after hearing that a neighbour’s child developed a fever, alongside her husband’s doubts and concern that her child might become fussy. Peers portraying caregivers received the same written brief and response prompts; they were not trained as standardized patients.

Data Collection and Instruments

Level 1 (Reaction): A 20-item, five-point Likert questionnaire assessed training content and role-play, the facilitator, training implementation, and workplace applicability. Five open-ended questions collected descriptive feedback. Internal consistency was high (Cronbach’s α = 0.905). The questionnaire was administered immediately post-training.

Level 2a (Knowledge): A 20-item multiple-choice test assessed GID principles, IPC, cultural sensitivity, vaccine hesitancy, and shared decision-making. Correct responses received 1 point, and incorrect responses received 0 points; the total score was calculated as (correct responses/20) × 100. The validated blueprint comprised five Greet, five Invite, and ten Discuss items, with raw maxima of 5, 5, and 10. Discuss contained more items because its blueprint covered broader content, but all items retained equal one-point weighting. Domain percentage was calculated as (correct responses in the domain/domain items) × 100. Two medical education and clinical communication specialists completed two rounds of content validation. Corrected item–total correlations ranged from 0.310 to 0.645, and Cronbach’s α was 0.846. The same test was administered immediately before the core training and during the post-training assessment beginning at 14:30.

Level 2b (Communication Skills): Skills were assessed once, immediately after the core training, through video-recorded peer simulations conducted from 14:30 in five parallel rooms. A pre-training video assessment was infeasible within the one-day field schedule and recording logistics; therefore, the outcome represents immediate post-training achievement, not improvement, retention, or workplace behaviour.

Three trained observers independently rated each recording using a 22-item checklist: Greet (6 items; maximum 12), Invite (8 items; maximum 16), and Discuss (8 items; maximum 16), giving a maximum total score of 44. Item-specific behavioural anchors defined 0 as not demonstrated, 1 as attempted but incomplete, unclear, or inconsistent, and 2 as correctly and completely demonstrated. For example, a fully performed open-ended-question item required a genuinely open question eliciting the caregiver’s perspective; a partially open question or one without appropriate follow-up received 1.

We averaged the three observers’ scores for each participant. Domain achievement was calculated as (mean domain score/domain maximum) × 100, and overall achievement as (mean total score/44) × 100. We interpreted these percentages descriptively, not as evidence of improvement or mastery.

Observer Training and Calibration

The three observers had backgrounds in medicine and medical or health professions education. They received structured instruction on checklist domains, behavioural anchors, confidentiality, and scoring only observable behaviour. Calibration used five practice videos excluded from the 24-recording analytic sample. Average measures, two-way mixed-effects, absolute-agreement ICCs were 0.809 overall, 0.868 for Greet, 0.598 for Invite, and 0.842 for Discuss. Given the five-video sample, these estimates guided calibration rather than providing definitive evidence of reliability. Because Invite agreement was lowest, observers clarified its more inferential items in a consensus meeting, then independently rated all study recordings, with no more than three recordings per observer per day to reduce fatigue.

Data Analysis

Frequencies, percentages, means, standard deviations, medians, interquartile ranges, 95% confidence intervals, and ranges were reported as appropriate. Mean-based 95% CIs for Level 1 and Level 2b scores used the t-distribution (df=23). We reviewed open-ended responses descriptively to contextualize quantitative findings; we did not conduct formal qualitative analysis or triangulation.

Knowledge scores were non-normally distributed (Shapiro–Wilk pre-test: W=0.877, p=0.007; post-test: W=0.834, p=0.001); therefore, we summarized them primarily using medians and interquartile ranges and analysed them using the Wilcoxon signed-rank test. Effect size was r=Z/√N, with N=24 paired observations. All observations were retained, individual change directions were tabulated, and potential ceiling restriction was examined from the post-test distribution and proximity to the scale maximum. We also performed paired analyses for each GID domain.

Inter-rater reliability for total and domain skills scores was assessed with 95% CIs using an average-measures, two-way mixed-effects, absolute-agreement ICC [ICC(A,3)], reflecting the mean ratings of three observers. Following Koo and Li (15), point estimates <0.50, 0.50 to <0.75, 0.75–0.90, and >0.90 indicated poor, moderate, good, and excellent reliability, while CI width informed precision. We treated achievement estimates with ICCs <0.75 as descriptive and hypothesis-generating. We also calculated exploratory post-hoc item-level ICCs, particularly for Invite, without replacing the primary total- and domain-level analyses. Associations between knowledge change and total and domain skills achievement were examined using Spearman’s rank correlation. Tests were two-sided with p<0.05, using IBM SPSS Statistics version 25 and Microsoft Excel.

Ethical Considerations

The Medical and Health Research Ethics Committee, Faculty of Medicine, Public Health, and Nursing, and Universitas Gadjah Mada approved the study (No. KE/FK/0302/EC/2026; 13 February 2026), and the Tana Toraja District Health Office granted administrative authorization. Participants provided written informed consent covering all procedures, including video recording.

To protect psychological well-being, we presented peer simulations as supportive learning evaluations rather than employment appraisals, and participants were informed that scores would not affect employment or institutional evaluations. The research team pseudonymized data and recordings, stored them on password-protected media accessible only to the research team and designated observers, and excluded them from employment or institutional reports.

Results

Participants’ Characteristics

All 24 participants completed the training and evaluation components (100% retention), representing all facilities in the sampling frame: 23 Puskesmas and one private maternal and child hospital. Eighteen were women (75.0%); 13 (54.2%) were aged 30–39 years, and eight (33.3%) were aged 40–49 years. Their mean age was 39.25 years (range 27–58). Sixteen participants were nurses (66.7%), and eight were midwives (33.3%); none had previously attended structured communication training.

Level 1: Participants’ Reaction

The mean overall reaction score was 4.66±0.29 (95% CI 4.54–4.78; scale 1–5; Table 2). All components scored above 4.00; training content and role-play received the highest rating (4.70±0.36), and training implementation had the lowest rate (4.60±0.35).

Recurring comments in the open-ended responses included appreciation for role-play, direct practice, and engaging facilitation. Participants requested improved pacing, more Invite practice, and greater room capacity, and considered GID relevant to daily immunization practice, particularly for parental hesitancy. Reported barriers to sustained application included time constraints, high caseloads, language diversity, and variable community receptiveness. We reviewed these responses descriptively rather than through formal qualitative analysis.

Level 2a: Knowledge Change

Knowledge scores were non-normally distributed (Shapiro–Wilk pre-test: W=0.877, p=0.007; post-test: W=0.834, p=0.001). Median scores increased from 77.50 (IQR 45.00–90.00) to 95.00 (IQR 85.00–100.00); corresponding means were 67.71±22.51 and 92.29±8.21. The Wilcoxon signed-rank test indicated a significant difference (Z=−4.023, p<0.001, r=0.82), with a mean change of 24.58±20.53 points (median 22.50; IQR 8.75–36.25).

Twenty-one participants had higher post-test scores, two decreased by five points, and one was unchanged. Scores clustered near the maximum, indicating a ceiling effect and restricted post-test variability; therefore, the effect-size estimate should be interpreted cautiously. All three GID knowledge domains also showed significant pre-to-post differences (all p≤0.001; Table 3).

Level 2b: Communication Skills Achievement

Because skills were assessed only once after training, these findings represent immediate post-training achievement rather than improvement or retention. Mean overall achievement was 67.80% of the checklist maximum (95% CI 60.1–75.5%; Table 4), with the highest achievement for Greet (80.21%), followed by Discuss (66.58%) and Invite (59.72%).

Average-measures inter-rater reliability was good overall (ICC=0.831), excellent for Greet (ICC=0.936) and Discuss (ICC=0.927), and moderate for Invite (ICC=0.574). The Invite estimate was lower than those of the other domains and had a wide 95% CI (−0.020 to 0.838); consequently, its achievement percentage should be treated as descriptive and hypothesis-generating rather than precise.

Item-level achievement was highest for eye contact and body posture (89.58%), partnership posture (88.89%), and vocal quality (88.89%) in Greet and adapting explanations (81.94%) in Discuss. Teach-back and exploring information sources had the lowest achievement (49.31% each), followed by exploring reasons for delay (51.39%) and paraphrasing or mirroring (52.08%; Table 5).

The exploratory post-hoc analysis indicated heterogeneous reliability within Invite. Active listening had poor reliability (ICC=0.288) despite relatively high achievement, whereas paraphrasing or mirroring had good reliability (ICC=0.832) despite low achievement. Exploring concerns (ICC=0.745) and exploring information sources (ICC=0.718) had moderate reliability. Thus, lower agreement was not uniform across Invite items and may have been concentrated partly in active listening. Given the post-hoc analysis and sample of 24, these findings are exploratory.

Knowledge change was not significantly associated with total skills achievement (Spearman’s ρ=−0.090, p=0.675) or with Greet (ρ= 0.071, p=0.741), Invite (ρ=−0.221, p=0.300), or Discuss achievement (ρ=−0.087, p=0.687). Given the small sample and restricted post-test knowledge variability, the absence of statistical significance should not be interpreted as evidence that knowledge change and skills achievement are unrelated.

Discussion

This study evaluated a one-day GID-based IPC training program for immunization officers in Tana Toraja, which reported South Sulawesi’s lowest full basic immunization coverage in 2024. Participants reported positive immediate reactions; knowledge scores were significantly higher after training, and the post-only assessment described immediate communication skills achievement. However, the single-group design precludes causal attribution, the absence of a skills baseline prevents estimation of improvement, and the lower reliability of Invite limits domain-specific interpretation.

The mean reaction score of 4.66/5.00 is consistent with evidence that adult learners value practical, participatory activities relevant to their work (16-18). Participants particularly appreciated role-play and direct practice, supporting the importance of experiential learning, feedback, and reflection in communication training (19). The relatively lower training-implementation rating and requests for more practice suggest that future sessions need improved pacing and additional guided Invite practice. Nevertheless, immediate reactions indicate acceptability, not learning, retention, or workplace effectiveness.

Median knowledge scores increased from 77.50 to 95.00, with a Wilcoxon effect size of r=0.82, comparable to pre-to-post differences reported after communication- and immunization-related training in Indonesia and Ghana (20, 21). Differences occurred across all GID domains, including the conceptually broader Discuss domain. However, post-test scores clustered near the maximum (median 95.00; IQR 85.00–100.00), restricting variability and potentially distorting the effect-size magnitude. Thus, r=0.82 should not be interpreted as an unbiased estimate of training effectiveness. Twenty-one scores increased, two decreased by five points, and one was unchanged; the study was not designed to explain these individual differences.

Immediate skills achievement was highest for Greet (80.21%), followed by Discuss (66.58%) and Invite (59.72%). Observable Greet behaviours, such as eye contact, posture, and vocal quality, can be modelled directly during a brief workshop, whereas Invite requires pausing, active listening, and exploration of concerns before offering advice. Such interactional skills generally require repeated practice, feedback, and varied scenarios (22-24). Each GID stage received approximately 30 minutes of role-play; this provided introductory exposure, but its adequacy for complex Invite behaviours was not established. Future programs should test repeated scenarios, feedback cycles, and follow-up practice rather than assume that one workshop establishes mastery.

The higher achievement in Discuss than in Invite may appear counterintuitive because, within the GID sequence, discussion should be informed by the caregiver’s perspectives and concerns elicited during Invite. One plausible explanation is that participants’ routine work as immunization officers may have provided greater familiarity with explaining vaccine information than with eliciting concerns and negotiating decisions. This interpretation is consistent with relatively high achievement in adapting explanations (81.94%) and explaining benefits and risks (77.78%). In contrast, teach-back (49.31%), involving caregivers in decision-making (69.44%), and respecting caregivers’ decisions (68.75%) were less consistently demonstrated. Discuss performance might, therefore, have reflected familiar information-provision behaviours more than consistently implemented shared decision-making, with some participants potentially proceeding to explanation without first sufficiently exploring caregivers’ perspectives. Because pre-training communication skills and the mechanisms underlying this pattern were not assessed, this explanation remains hypothesis-generating.

Teach-back and exploring information sources had the lowest item-level achievement (49.31% each). Teach-back supports assessment of understanding and clearer communication (25), while information sources are relevant because vaccination decisions may be shaped by family, community narratives, social media, and previous experiences (26). These scores may reflect limited familiarity, insufficient practice, scenario characteristics, or culturally shaped interaction. In Tana Toraja’s clan-based, community-oriented context, exploring private beliefs and information sources may require locally appropriate wording and relationship-building. Because these mechanisms were not tested, they remain hypotheses rather than established explanations.

The Invite-domain ICC point estimate was moderate (ICC=0.574) under the specified thresholds (14), but it was lower than those for Greet and Discuss and had a wide 95% CI (−0.020 to 0.838). Consequently, the 59.72% achievement estimate is imprecise and may partly reflect inter-rater measurement error rather than actual performance; it should be treated as hypothesis-generating, not definitive evidence of poorer competence. Post-hoc findings suggested that uncertainty was not uniform: active listening had poor reliability (ICC=0.288) despite relatively high achievement, whereas paraphrasing or mirroring and exploring information sources showed higher reliability despite low achievement. However, the small sample and exploratory analysis limit firm conclusions.

Several factors may have reduced agreement. Invite items require judgments across an interaction rather than identification of a discrete action, which is challenging in video-based assessment (27). Peers rather than trained standardized patients portrayed caregivers; portrayal consistency was not formally evaluated, and calibration used only five videos. Before definitive domain-level interpretation, the Invite checklist requires revision and revalidation with clearer behavioural anchors, more extensive calibration, and more standardized caregiver portrayals.

Knowledge change was not significantly associated with total or domain-specific skills achievement. Given the small sample and restricted post-test knowledge variability, this should not be interpreted as evidence that knowledge and skills are unrelated. Although scores from different instruments cannot be compared directly, near-ceiling knowledge scores alongside lower skills achievement are consistent with the distinction between acquiring declarative knowledge and translating it into observable performance. Communication behaviours require deliberate practice, feedback, and repetition before becoming fluent and consistent (22-24). The training may, therefore, have supported conceptual understanding without providing sufficient practice for stable performance; habit formation and workplace transfer require longitudinal assessment at Kirkpatrick Level 3.

Communication training should be one component of a broader immunization strategy. Tana Toraja’s 43.06% coverage cannot be attributed to communication alone and requires concurrent attention to vaccine supply, service access, outreach, data quality, and community engagement. GID offers a culturally informed starting point because it was developed for hierarchical, community-oriented relationships in Indonesia and Southeast Asia (8, 9). Still, its transferability to Tana Toraja has not been formally validated.

Previous GID research linked training with improved doctor–patient communication and clinical outcomes in chronic-disease care (9). This study extends GID toward preventive service delivery, specifically immunization as specific protection within primary prevention. However, chronic-disease findings cannot be extrapolated directly, and the pathway from GID training to caregiver trust or vaccine uptake remains theoretical. Stronger studies using patient-, behavioural-, and coverage-level outcomes are required. This first reported district-level application in Tana Toraja, therefore, provides preliminary information for program refinement rather than evidence of effectiveness.

Reported barriers, including time constraints, high caseloads, language diversity, and variable community receptiveness, require practical responses. A simplified bilingual Indonesian–Torajan GID job aid could be co-developed and culturally tested with local health workers and community representatives. Brief practice, case discussions, peer coaching, and checklist-based feedback could be integrated into existing monthly supervision meetings to avoid increasing workload. Collaboration with community health workers and influential family, clan, community, or religious figures may improve cultural adaptation and receptiveness.

This evaluation involved 24 non-randomly selected representatives from one district, and one participant per facility could not capture within-facility variation. The single-group design, immediate knowledge testing, ceiling effect, and greater representation of Discuss items limited causal, effect-size, and retention inferences. Post-only skills assessment measured achievement rather than improvement, retention, or workplace behaviour. Peer caregivers, five-video calibration, lower and imprecise Invite reliability, and an instructional dose not validated for complex skills further limited interpretation. Reaction ratings remained susceptible to social desirability despite confidentiality and non-employment assurances, while open-ended responses underwent descriptive rather than formal qualitative analysis. Future studies should evaluate Kirkpatrick Level 3 through observed workplace GID use and Level 4 through vaccine acceptance, uptake, missed opportunities, and coverage, using baseline and follow-up measurements, comparison groups where feasible, and adjustment for concurrent system interventions.

Conclusion

The one-day GID-based IPC training was well received and was associated with a significant immediate pre-to-post increase in knowledge across all three GID domains. However, the single-group design, immediate assessment, and post-test ceiling effect preclude causal attribution and limit interpretation of effect magnitude and retention. Overall post-training skills achievement was 67.80%; because no pre-training skills baseline was collected, this finding represents immediate achievement rather than improvement. Achievement was highest for Greet and lowest for Invite although the Invite estimate should be treated as hypothesis-generating because its moderate inter-rater reliability was lower than that of the other domains and limits precise interpretation. Future training should provide repeated practice in teach-back, paraphrasing, and exploring caregivers’ information sources and reasons for hesitancy, while revising the Invite checklist and strengthening observer calibration. These findings do not establish retention, workplace behavioural transfer, or improved vaccine uptake. Longitudinal comparative studies incorporating baseline skills assessment should evaluate workplace behaviour at Kirkpatrick Level 3 and immunization outcomes at Level 4.

Acknowledgements

The authors thank the Tana Toraja District Health Office and the leadership and staff of all 23 Puskesmas and RS Fatima for their institutional support and willingness to participate in this study. The authors also thank Portkesmas for partial funding support, technical coordination, and field documentation during the training and data collection activities. The authors are grateful to the three observers for their rigorous and independent contribution to the video-based communication skills assessment. This study was conducted as part of the requirements for a master’s degree in Medical and Health Professions Education. The support provided by Portkesmas did not influence the study design, data analysis, interpretation of findings, or manuscript preparation.

Conflict of Interest

The authors declare no personal financial conflicts of interest related to this study. One author serves as Program Specialist at Portkesmas, which provided partial funding, technical coordination, and field documentation support for the training activities described in this manuscript. This relationship was disclosed, and the support from Portkesmas did not influence the study design, data analysis, interpretation of the results, manuscript preparation, or the decision to submit this manuscript. One author was involved in the earlier development of the Greet–Invite–Discuss (GID) Guideline. This was disclosed, and the evaluation data were analysed and reported independently.

Declaration of AI

Artificial intelligence-based language tools were utilized to facilitate translation, improve textual structure, and increase clarity during the preparation and editing of this manuscript. These tools were not involved in study conception, data collection, statistical analysis, data interpretation, or the generation of research findings. All outputs produced by these tools were comprehensively reviewed, critically revised, and approved by the authors, who assume full responsibility for the manuscript’s content and accuracy.

Authors’ Contribution

MA.I conceived and designed the study, prepared and coordinated the training, collected and analyzed the data, interpreted the findings, and drafted the manuscript. TS.P contributed to study supervision, evaluation design, methodological refinement, interpretation of findings, and critical revision of the manuscript. M.C contributed to the GID framework, study supervision, methodological refinement, interpretation of findings, and critical revision of the manuscript. All authors approved the final manuscript and agreed to be accountable for the work.

References

1. World Health Organization. Immunization coverage [Internet]. Geneva: World Health Organization. 2024 [Cited 2026 Jan 10]. Available from: https://www.who.int/news-room/fact-sheets/detail/immunization-coverage
2. Kementerian Kesehatan RI. Strategi Komunikasi Nasional Imunisasi 2022–2025 [Internet]. Jakarta: Kementerian Kesehatan RI. 2022 [Cited 2026 Jan 10]. Available from: https://kemkes.go.id/app_asset/file_content_download/16653827576343b965228c40.04885132.pdf
3. Dinas Kesehatan Provinsi Sulawesi Selatan. Laporan Cakupan Imunisasi Dasar Lengkap Sulawesi Selatan 2024 [internal report]. Makassar: Dinas Kesehatan Provinsi Sulawesi Selatan; 2024.
4. Basheer SA. Improving routine immunization coverage through strengthening interpersonal communication skills of primary health care workers in rural local government areas in Kebbi State, Nigeria [dissertation]. Bangkok: Chulalongkorn University; 2020.
5. Mahanta TG, Nirmolia N, Biswas AB, Barua SD. Service delivery in postpandemic period for routine childhood immunization in Assam following capacity building with BRIDGE IPC Skill tool. J Fam Med Prim Care. 2024; 13(10):4482-5.
6. UNICEF. Interpersonal communication for immunization: training for front line workers [Internet]. Geneva: UNICEF; 2019 [Cited 2026 Jan 10]. Available from: https://www.unicef.org/eca/media/8576/file/interpersonal-communication-participant.pdf
7. Nicol E, Turawa E, Bonsu G. Pre- and in-service training of health care workers on immunization data management in LMICs: a scoping review. Hum Resour Health. 2019; 17:92.
8. Claramita M, Tuah R, Riskione P, Prabandari YS, Effendy C. Comparison of communication skills between trained and untrained students using a culturally sensitive nurse-client communication guideline in Indonesia. Nurse Educ Today. 2016; 36:236-41.
9. Claramita M, Arininta N, Fathonah Y, Kartika S, Prabandari YS, Pramantara IDP. A partnership-oriented and culturally-sensitive communication style of doctors can impact the health outcomes of patients with chronic illnesses in Indonesia. Patient Educ Couns. 2020; 103(2):292-300.
10. Silverman J, Kurtz S, Draper J. Skills for communicating with patients. Boca Raton: CRC Press; 2013.
11. Kirkpatrick JD, Kirkpatrick WK. Kirkpatrick’s Four Levels of Training Evaluation. Alexandria: Association for Talent Development; 2016.
12. Frye AW, Hemmer PA. Program evaluation models and related theories: AMEE Guide No. 67. Med Teach. 2012; 34(5):e288-99.
13. Gagné RM, Wager WW, Golas KC, Keller JM. Principles of instructional design. Belmont: Wadsworth/Thomson Learning; 2005.
14. Iqra MA. Evaluasi Pelatihan Komunikasi Antarpribadi (KAP) Berbasis Greet-Invite-Discuss (GID) Guideline pada Petugas Imunisasi di Kabupaten Tana Toraja [master's thesis]. Yogyakarta: Universitas Gadjah Mada; 2026.
15. Koo TK, Li MY. A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med. 2016; 15(2):155-63.
16. Taylor DCM, Hamdy H. Adult learning theories: implications for learning and teaching in medical education: AMEE Guide No. 83. Med Teach. 2013; 35(11):e1561-72.
17. Traicoff D, Tchoualeu D, Opare J, Wardle M, Quaye P, Sandhu HS. Applying adult learning best practices to design immunization training for health care workers in Ghana. Glob Health Sci Pract. 2021; 9(3):487-97.
18. Knowles MS, Holton EF, Swanson RA. The adult learner: the definitive classic in adult education and human resource development. New York: Routledge; 2020.
19. Byrne M, Campos C, Daly S, Lok B, Miles A. The current state of empathy, compassion and person-centred communication training in healthcare: an umbrella review. Patient Educ Couns. 2024; 119:108063.
20. Rachmah Q, Rachmayanti RD, Rochmah TN, Devy SR, Andari S, Ismayani The efficacy of interpersonal communication training for health workers in behavior change related to stunting. Media Gizi Kesmas. 2023; 12(1):410-6.
21. Tchoualeu DD, Harvey B, Nyaku M, Opare J, Traicoff D, Bonsu G. Evaluation of the impact of immunization second year of life training interventions on health care workers in Ghana. Glob Health Sci Pract. 2021; 9(3):498-507.
22. Denniston C, Molloy E, Nestel D, Woodward-Kron R, Keating JL. Learning outcomes for communication skills across the health professions: a systematic literature review and qualitative synthesis. BMJ Open. 2017; 7(4):e014570.
23. Holcomb J, Ferguson GM, Thornton L, Highfield L. Development of training curriculum to improve patient communication skills and social support among community health workers. Health Lit Res Pract. 2022; 6(2):e142-50.
24. Peimani M, Tanhapour M, Majlesi M, Nasli-Esfahani E. Training in communication skills for healthcare providers in chronic care: a systematic review. BMC Med Educ. 2025; 25:1314.
25. Talevski J, Wong Shee A, Rasmussen B, Kemp G, Beauchamp A. Teach-back: a systematic review of implementation and impacts. PLoS One. 2020; 15(4):e0231350.
26. Di Lorenzo A, Stefanizzi P, Tafuri S. Are we saying it right? Communication strategies for fighting vaccine hesitancy. Front Public Health. 2024; 11:1323394.
27. Yeates P, Moult A, Lefroy J, Walsh-House J, McKinley R, Fuller R. Understanding and developing procedures for video-based assessment in medical education. Med Teach. 2020; 42(11):1250-60.