Effect of Integrating Traditional Tuberculosis Care With Modern Health Care on Case Detection

Tuberculosis Clinical Trial

Official title: Effect of Integrating Traditional Tuberculosis Care With Modern Health Care on Case Detection, Cost, and Client Satisfaction in Amhara Region, Ethiopia

Status Recruiting

Clinical Trial Summary

Although many interventions are implemented to increase TB case detection, decrease diagnosis delay, and avoid catastrophic costs, there are no significant changes and the end TB goal will not be achieved in 2035. Innovative intervention that considers indigenous knowledge and unique culture and religious perspectives because many people go to traditional healers and holy water for healing. Therefore, integrating traditional tuberculosis care with modern care increase case detection, decrease diagnosis delay, and avoid catastrophic costs. There is no literature clearly defining integrating traditional TB care with modern care, but for the purpose of this study, integrating traditional care with modern care is defined as the collaboration of two systems through referral linkage. TB screening and diagnosis services will be done collaboratively in traditional and modern care services. A referral linkage model will be used to detect TB cases in both traditional and modern care services. Health care providers, traditional healers, priests, pastors, and imams will participate in the integration process. TB detection or diagnosis services will be integrated through referral linkage and strengthening capacity-building strategies. Traditional care centers and modern health care services will work collaboratively to improve TB case detection, reduce care costs, and avoid diagnosis delays. The standardized operational procedure of the full interventional package is described below.

There are four steps of the intervention phases. These are the preliminary phase, preparation for implementation and refinement on a small scale phase, administering the intervention, and end-line assessment of outcomes. The intervention will be providing training for traditional and modern care practitioners, patient education, TB screening, and bidirectional referral linkage.

This study hypothesized that integrating traditional care with modern care at the primary care level will increase the TB case detection rate by fifteen percentage points. Integrating traditional care with modern care at the primary care level will decrease TB diagnosis delay by fifteen percentage points. Integrating traditional care with modern care at the primary care level also will decrease the cost of TB care by 15 percentages of points

Clinical Trial Description

Quantitative description Design: Both qualitative and quantitative methods will be used to answer our research questions. Cluster randomized control trial study design will be used to determine the effectiveness of integrating traditional care into modern care on case detection rate (for the first, second, and third objectives). The explanatory sequential design will be used to determine the effect of integrating traditional care and modern care on patient satisfaction (fourth objective). This explanatory sequential design will follow to obtain substantiate evidence from quantitative data which is used to determine whether it has effects on patient satisfaction on integration between the traditional with modern care. In this design, two phases will be involved. The first phase will be an initial quantitative instrument phase, followed by a qualitative data collection phase, in which the qualitative phase builds directly on the results from the quantitative phase. In this way, the quantitative results will be explained in more detail through the qualitative data.

Study participants: Study participants will be TB suspected individuals (i.e., coughs for more than two weeks, and other constitutional TB signs and symptoms such as fever, chills, night sweats, anorexia, and weight loss).

Intervention: Integration is recognized as a multidimensional concept with no singular purpose. Integrated service delivery can be defined as joining inputs, management, organization, and distribution of health services related to diagnostics, treatment, care, rehabilitation, and health promotion as a means of improving health status, access, quality, and continuity of care, consumer satisfaction, and efficiency. There is no literature clearly defining integrating traditional TB care with modern care, but for this study, integrating traditional care with modern care is defined as the collaboration of two systems through referral linkage. TB screening and diagnosis services will be done collaboratively in traditional and modern care services. A referral linkage model will be used to detect TB cases in both traditional and modern care services. Health care providers, traditional healers, priests, pastors, and imams will participate in the integration process. TB detection or diagnosis services will be integrated through referral linkage and strengthening capacity-building strategies. Traditional care centers and modern health care services will work collaboratively to improve TB case detection, reduce care costs, and avoid diagnosis delays. The standardized operational procedure of the full interventional package is described below.

Phase 1: Preliminary phase A preliminary intervention package is developed by inviting TB experts, and language professionals. The training manual is prepared and will deliver for practitioners to increase their knowledge and skill. Based on the provision of training contents, the knowledge gain and skill acquisition will be evaluated. The fidelity of the intervention is developed to monitor the whole implementation activities of the training manual as planned.

Phase 2: Preparation for implementation and refine on a small scale In this phase, numerous activities related to the intervention will be performed. Experts who can implement the intervention will be recruited based on experts' educational status (who have earned a BSc degree and above in nursing or public health or medical doctor), and have research and working experiences at TB clinic. A pilot test will be implemented outside the study area on 10% of the project sample size which is suggested by Connelly LM (2008)(287). Based on pilot findings and experts' judgment, the intervention package will be refined and enriched. Before the actual intervention is administered, a baseline data collection will be conducted on TB case detection rate and diagnostic delay in both intervention and control sites.

Phase 3: Administering the intervention, feedback, and revision This phase is crucial that the researchers are planning to administer intervention successfully. In this phase, a full-blown intervention implementation will be conducted in the intervention areas. Therefore, practitioners will screen all TB suspected cases at traditional care centers, and every suspected case will be referred to modern TB care centers for further diagnosis and treatment. Along with the administration of this intervention, the feasibility and acceptability of the intervention will be assessed using standardized checklists (see annexed checklist).

Phase 4: End line assessment of outcomes After the completion of the intervention period, end-line data collection will be conducted on both intervention and control groups. The outcomes of the intervention will be evaluated quarterly, half-year, and one year by trained data collectors. The comparison will be done between the end-line and the baseline data in both intervention and control groups.

Training on basic concepts of TB, causes of TB, signs, and symptoms, mode of transmission, screening, diagnosis, case detection, treatment, the benefit of early detection, complication, and prevention will be delivered to both traditional and modern care practitioners. The training will also include topics about intergrading traditional and modern health care systems. The training manual is prepared and annexed in this document (see annex 7: Training manual). The training will be delivered at Bahir Dar University College of Medicine and Health Sciences meeting hall. The training will be provided three times which will be delivered every three months. In the first round, traditional practitioners will be trained for 5 days. In this round, the training will be also delivered to modern health care providers on how to communicate and work with traditional care providers for 2-day. Then, a one-day interactive discussion session will be prepared for both traditional and modern care providers. In the second round and third rounds, one-day training will be carried out at three and six months. Training of the Trainer (TOT) will be given by researchers and those experts who had TOT will provide training for traditional care providers (traditional healers, priests, Imams, and pastors) and modern care providers (nurses, public health officers, physicians, laboratory technicians/ laboratory technologists).

Furthermore, researchers will recruit traditional care providers (TCP) (traditional healers, priests, pastors, and imams), and modern care providers (MCP) who can implement the intervention. The traditional healers, priests, pastors, and imams will take a brief history of patients' illnesses and complete the documentation properly. The suspected patients will be referred to the nearby health facilities in the intervention district that fulfill the inclusion criteria. Then, trained healthcare providers who are working in the selected health facilities will screen and diagnose patients based on the national TB treatment guideline and diagnostic algorithm. The intervention, content, duration, frequency, and dose are presented in table 1 below.

The intervention of fidelity (IF) refers to the degree to which an intervention is delivered as intended by researchers. The intervention fidelity is developed based on behavioral change consortium treatment fidelity recommendations. The fidelity checklist is developed. This fidelity will be used to monitor whether the intervention is applied exactly as we have planned. The checklist depends on adherence, quality intervention delivery, dosage, participant responsiveness, and compared between experimental and comparison groups. To select the trainer, the authors will be select the trainer score more than 80% pass mark on post-test to each content. The qualification of the trainee (practitioners) will be considered after scoring more than 60% post-test to each content of the course. The content of the training will be assured by recording and supervising each session (see annex table 3). The patient will be referred from traditional care centers to modern care centers using the TB screening and referral form included in the annex (see TB screening and referral form).

The control sites will be followed the existing passive case-findings system (self-referral patients to nearby health facilities that use the same national guidelines to treat TB). The baseline information will be assessed in both intervention and control groups at the same time. The findings obtained from the control groups will be compared with the intervention groups. Finally, changes among the two groups will be assessed and concluded.

Sample size determination: To estimate the effect of integrating traditional care with modern care on TB case detection, cost of TB care, diagnosis delay, and patient satisfaction, the sample size is calculated by using the two-sample comparison of proportions formula using STATA version 16. Fisher's exact test for two independent groups is used to calculate the final sample size.

By considering, the previous study conducted in the Amhara region with the proportion of 59.9% and 0.05 two-tailed alpha, power 80% by assuming that the effect of the intervention on diagnosis delay decreased by 15% in the intervention group as compared to the control. Adding 10% contingency, the final sample size is 814 (407 in each arm). Since the study is a multicenter individually randomized trial, a design effect is used. According to the multicenter design effect formula, design effect formula is design effect=(σe^2)/(σe^2+σB^2 ) = 1-ρ. The majority of formulas applied on multicenter design with better accuracy relative difference (ICC lesser than 0.052) for designs including more than 10 centers. Assuming that randomized is balanced and stratified on centers that have equal group sizes. Assuming the variation between the centers is 0.05, the design effect = 1-ρ=0.95. Therefore, the sample size is taken as the final sample size is calculated as sample size* design effect 814*0.95=774 (387 per group).

Participant recruitment and randomization: Randomization is stratified by the district. Based on the list obtained from the zone, two Woredas and two town administrations will be selected by random sampling methods using a random.org website. Of these, one district and one town administration were assigned in the intervention group, and another district and town administration were assigned in the control group. All health facilities and traditional care centers will be allocated to intervention or control arms using random.org. Allocation will be carried out by a trial statistician before the intervention or patient recruitment. Using the rule of thumb assumption, about 20% of districts in the South Gondar Zone will be selected randomly for this study. For the cluster, randomized control trial, the intervention, and control districts have been randomly selected. Based on the random sampling approach Dera woreda and Debre Tabor were selected as intervention sites and Libokemkem woreda and Woreta town were selected as control sites. The intervention and control sites have almost similar baseline TB case detection rates. In addition, health facilities to population ratio, traditional care providers, and socio-culturally aspects of both intervention and control sites are almost similar. The random allocation process, enrolling participants and participants assigned in the intervention group will be done by statisticians who will not have awareness about the characteristics of the intervention and control groups using the computer random.org website. The study subjects will be unaware of the intervention they received. Laboratory personnel who will be involved in the tuberculosis diagnosis and a person who will assess the outcomes of intervention will be unaware of the group assignment. Intervention outcomes will be assessed by research assistants who are not participating in participant recruitment and they are masked. The study participants, caretakers/attendants, and healthcare providers will be masked at any point.

Trial safety and adverse effects monitoring: Integrating traditional care with modern care is the trial-tested by this study. This trial will not directly involve any procedures or drugs that affect the study participants and there has no known risk of any adverse event. A multi-disciplinary group of specialists i.e. Data Monitoring Committee (DMC) will be formed that makes recommendations to those in charge of the overall management of the trial. Also, interim analyses will be conducted based on the pre-specified trial protocol to minimize bias from unplanned analyses. DMC recommendations will be included either: continuing the trial as planned; stopping early for hazard; stopping because efficacy is unequivocally established, or stopping because continuing the trial is futile.

Trial adherence and compliance monitoring: Ongoing process evaluation of the intervention will be carried out. Regular monitoring of intervention and control groups will be taken place by trained assessors. Overall adherence of intervention and control of study participants in trials will be assessed using self-report and direct observation by trained field supervisors.

Participant retention strategies: Once TB suspected individuals are enrolled in this study, trained traditional healers and clergy will provide health promotion and avocation to the patients who go to modern care centers (health facilities) with referral letters. Regular communication and feedback will be made between the supervisors and modern and traditional care practitioners to minimize the attrition rate of intervention participants from the study.

Participant withdrawal criteria: At any time, a subject can voluntarily withdraw from participation in some or all components of a study for which the subject previously consented to participate, the researchers may terminate a subject from some or all study components at any time at his or her discretion. The subject will be terminated from the study by the investigators that are believed the subjects' safety is compromised or feeling of discontinuation described by the subject. When the research team decided to withdraw the subject from the study, the reason will be informed to them. The data will be handled by obtaining additional identifiable private information from any source such as medical records, or any laboratory samples that have not yet been analyzed. In addition, retaining and analyzing previously collected data.

Study variables: Outcome variables of this study are case detection rate, diagnostic delay, cost of TB care, and client satisfaction. On the other hand, the independent variables are socio-demographic variables such as gender, age, educational status, marital status, occupational status, religion and income and stigma related to TB infection, patients knowledge on TB, health-seeking practice, contact history, accessibility of facilities for TB diagnosis, transportation access, presence of treatment support and distance.

The outcome variables will be measured based on their outcome interest. Case detection rate will be calculated as the number of cases notified divided by the number of cases estimated for that year, expressed as a percentage. The WHO estimator for the year 2020 was 140 per 100,000. To estimate the number of TB cases for that year will be computed by using X/100,100* total population of the catchment area equal to Z. Where "X" is given by the WHO for that year, it is varied across the countries and "Z" is several cases estimated for that year. Thus, CDR=number of cases notified divided the number of cases estimated multiplied by 100 or (notified cases/Z)* 100, expressed as a percentage.

To assess patient satisfaction a short form patient satisfaction questionnaire (PSQ-18) which is a validated tool adapted from a previous study. In addition, the tool is adapted from patient satisfaction literature. Patient satisfaction with the latest visit will be assessed using five items on a five-point Likert scale ranging from strongly disagree (1) to strongly agree (5). The patient will be considered satisfied when the individual scores above the mean.

These are payments made directly by the patient or their household member. Direct medical payments will include payments for formal medical professionals, informal traditional or alternative practitioners, clinics, health centers, pharmacies, and hospitals for medical services and products. This will exclude prepayment for health services for example health insurance payments and where relevant, net of any reimbursements to the individual who made the payments. The other will be non-medical direct costs including travel, accommodation, food, or other non-medical payments that are incurred by the patient, their household member, or caregiver/guardian while picking up medication or during the visit/hospital stay for TB care. The sum of the direct medical and non-medical costs for the different phases/entire episode is used as the numerator.

Indirect costs can be estimated using two alternative methods: self-reported household income loss net of welfare or social assistance payments. This includes the cost that the caregivers bear by contributing their time and in-kind services. Any gain or loss of income reported during each phase or for the entire episode is considered the total indirect cost. The opportunity cost of time spent away from the daily productive routine. Participants are asked to self-report time spent seeking and receiving care as well as the average number of hours, they work each day if this changed since the last interview, and if so, by how much. The time total period of absence (in hours) is multiplied by the hourly wage rate of the absent worker. The hourly wage can be estimated from directly reported data, household asset ownership, or national statistics. Indirect costs for the different phases/entire episode will be used as the numerator in the catastrophic total cost equation. The indirect cost will include productivity loss of the patients and caregivers during TB care. The catastrophic cost will be a total cost of tuberculosis care (including the direct and indirect cost) exceeding 20% of the annual income of the household. Costs will describe using quartiles and expressed as the median and 25-75th interquartile range. Patients with costs >75th quartile will consider having high expenditure and costs <75th quartile will consider as a low-cost expenditure.

Sampling procedure and technique Study areas and subjects will be selected randomly. For the first objective, all forms of TB cases in the health facilities will be taken from patients in the health facilities. For the second, third, and fourth objectives a proportional allocation will be done for each health facility and participants will be selected randomly. All health facilities providing TB diagnosis and treatment service will be identified.

Data collection methods: A structured questionnaire will be adapted from previous literature. The structured questionnaire was composed of socio-demographic and socioeconomic, perceptions, patient knowledge, service accessibility, stigma, contact history, patients' satisfaction, cost of TB (catastrophic cost), diagnosis delay, and healthcare-seeking practices. The questionnaire will be translated into the national language (Amharic) spoken by almost all residents in the study area. The data will be collected using face-to-face interview questionnaires. The data will be collected by BSc nurses or public health officers who had data collection experiences. Nurses or public health officers who had MSc/MPH degrees will be selected for data collector supervisors. The data will be collected at baseline and the end of one year of the intervention.

Data quality assurance: Data collectors will be recruited from BSc nurses and health officers. The data collection will be supervised by experienced MSc nurses or MPH experts who have sufficient experience in supervision. Training which comprised of the aim, methodology, sampling technique, ethical issues, and data collection instrument and data collection procedure will be prepared and delivered for implementers, data collectors, and supervisors. The training will be done by researchers. Professional and language experts will invite to validate the tools, training guidelines, and intervention manuals. Then, the tool will be examined by conducting a pretest among 5% of study participants in non-selected woredas. The intervention will be conducted by experts who have certificates in TB training. The necessary amendment will take place based on the pretest findings. The researcher will be in regular contact with the data collectors with scheduled meetings and other communication channels as needed to maintain the quality of data. To minimize data entry errors, a double data entry method will be used. Buffering zone will be considered between the intervention and control areas to avoid information contamination. All potentially eligible populations in the selected areas will be included in the study. The primary outcome assessors will be unaware of the intervention. The primary assessor will be a trained TB expert who has no information about the intervention group and control group. Additional outcome measurements will be performed in the same manner in both arms. The data will be analyzed using intention-to-treat principles. To ensure the quality of methodological components, a Consolidated Standards of Reporting Trials 2010 statement extension to cluster randomized control trial study guideline will be used and to ensure better reporting of the intervention Template for Intervention Description and Replication (TIDieR) checklist will be used.

Data processing and management: In the field, data will be checked for consistency and completeness every day by supervisors and the principal investigator. Then, a data entry template will be created on EpiData version 3.1 based on coded responses. Data clerks will enter the data of which 10% will be randomly double-checked by the principal investigator. The data will be then exported to STATA version 16 where further recoding and cleaning will be done to check for inconsistencies, outliers, and missing values. Finally, the cleaned data will process and analyzed using STATA. Before further processing of the data, numeric data will be checked for normality using normality plots (Q-Q plots and/or histograms) or normality tests. Internal consistency (Cronbach's Alpha (α) =0.75) will be also checked for the reliability of the items.

Description analysis: The data will be described using frequency, proportions, mean, median, standard deviation, interquartile range, tables, and graphs. Comparisons of proportions among categorical variables will be made using the Chi-squared test. Also, associations between numeric variables will be examined using simple correlation analysis. For those variables which have a normal distribution, a comparison of mean among different categories will be made using independent t-test, paired t-test, and one-way Analysis of variance (ANOVA). For non-normal distributions, median across categories will be compared using a nonparametric test. In all the tests statistical significance will be judged at p<0.05. Generalized Estimating Equation (GEE) with a binary logit function will be used to examine the change difference of outcome between the intervention and control groups. GEE will be run to accommodate the correlation of observations within subjects. The correlation matrix test will be assessed and the fit model will be selected. The effects of potential confounding factors will be accounted for. The change difference between independent and dependent variables will be analyzed. The effect of the intervention on the TB case detection rate will be assessed at the end of one year. Odds ratios and respective 95% confidence intervals will be calculated and p values < 0.05 will be considered statistically significant.

Generalized linear mixed models (GLMM) will be used to assess the impact of the intervention on the change of cost of TB care over time and diagnosis delay. This GLMM enables to accommodate the correlation of observations due to the pre and post-intervention measurements and clustering individuals within randomly selected clusters. Model fitting will be checked before the analysis is done. The effects of potential confounding factors will be controlled. The variance of the cluster-level residual errors will be computed using the intercept-only model. The variation between and within the groups will be analyzed.

Patient satisfaction-related data will be analyzed using STATA software V.16.0. Owing to the ordinal nature of the outcome variable (low, medium, and high satisfaction), a typical approach will be used to use the standard ordered logit. The assumption will be checked. Independent variables will be checked as continuous, categorical, or ordinal. The Multi-collinearity test will be checked whether two or more independent variables are highly correlated with each other. In addition, proportional odds will test that each independent variable has an identical effect at each cumulative split of the ordinal dependent variable. Statistical significance for the final model was set at p<0.05.

Qualitative methods Regional, zonal, and district TB focal persons, traditional healers, religious leaders (clergy), Pastors and Imams as well as health care providers will be participated in in-depth interviews (IDI) and focus group discussions (FGDs). The purposive sampling method will be used to identify and select information-rich subjects for the most effective use of limited resources. Tentatively the sample size will be 60 (10 focuses group discussions (FGDs)) each group will contain 6 members), 50 In-Depth Interviews (IDI).

Qualitative data collection methods: To obtain more evidence that is detailed linked to the study, both FGD & in-depth interview data collection methods will be applied. Interview guiding questions will be developed from similar literature. Question guides for FGDs and IDI will be prepared based on the context of the study participants. The interview guide will be prepared in English. The IDI or FGDs will be conducted by the local language Amharic. The data will be collected by the researcher. All conversations will be captured by audiotape recorder. Data will be collected until saturation of response is assumed to be maintained. The saturation of the response will be decided based on the adequacy of information, completeness of the information, and little new information coming from the interviewees and discussants. The interview will take place at a convenient place and time. Ten FGDs each consist of at least six members. According to scholars' suggestions, 50 participants will be adequate for IDI. The IDI will be conducted among health workers, TB focal persons, traditional healers, Imams and clergy, and traditional healers. In addition, FGDs will be conducted among TB patients. The IDI and FGDs guiding questions will be checked by language experts, professionals, and senior consultants to maintain the quality of the tool. The guiding questions were translated into the official language, Amharic to easy understanding by the interviewees. Then, it will be translated back into the English language to maintain consistency. Before FGDs and IDI are conducted, quiet rooms or settings will be arranged in the field.

The data will be analyzed using sequences of steps. First, the data will be prepared and organized. The recorded interview will be transcribed all data from voice to text verbatim. Then, the data will be printed out the transcripts, gather your notes, and documents. Data transcription will be done by the researcher to assure the dependability of the transcription. The transcribed data will be checked for its consistency. When there is variation in transcription between the transcribers, the discussion will be made by the researchers. The transcribed data will be translated to the English language for the subsequent analysis. Secondly, the data will be reviewed and explored. Thirdly, after repeated reading, the researcher will create initial codes using highlighters, notes in the margins, sticky pads, concept maps, or anything else that helps the researcher to connect with the data. Fourthly, review those codes and revise or combine them into themes. Finally, themes will be presented cohesively. Thematic analysis will be used to describe the data obtained with interviews and focused group discussions. In vivo Software will be used for data analysis and the result will be explained after the consensus is reached.

Ethics: The study will be approved by an institutional review board (IRB) of Bahir Dar University, college of medicine and health sciences. Informed consent for adult participants and assent for pediatrics participants will be obtained before data collection started. The purpose of the study will be communicated and permission will be obtained from the respected officials; the regional health bureau, respected zonal, words health offices, and health facilities.

Participation of the study subjects will depend voluntarily. Hence, informed consent will be taken from the study subjects. The information sheet, adult informed consent, and written assent forms are prepared (Annex 1-3). The information included the purposes of the study, the rights of participants, potential benefits, and harms of the study. The study will not consider any financial compensation for participation. Participants will not enforce in any way by anybody to take part in the study. To ensure the confidentiality of the information provided by the participants, personal identifiers will not be included. In addition, questionnaires will be kept in a locked cabinet and data in a password-protected computer accessible to the principal investigators. The study will not have any harm to participants and effort will be made to maximize the benefit to the participants. ;

Related Conditions & MeSH terms

• Latent Tuberculosis
• Tuberculosis
• Tuberculosis Infection
• Tuberculosis, Pulmonary

• NCT number: NCT05236452
• Study type: Interventional
• Source: Bahir Dar University
• Contact: Desalegne Amare
• Phone: 0929466624
• Email: desalegnezelellw@gmail.com
• Status: Recruiting
• Phase: N/A
• Start date: July 1, 2022
• Completion date: December 30, 2023