The Determinants of Non-compliance on Rabies Vaccination in North-West Peninsular Malaysia

Introduction Rabies post-exposure vaccination (Rabies PEV) remains the most fundamental prevention of human Rabies if administered in a timely and appropriate manner. The study was aimed to determine the proportion and determinants of non-compliance on Rabies PEV among dog bite patients in Perlis, Malaysia from July 2015 to June 2020. Methods A retrospective cohort study was conducted using Perlis Rabid Potential Animal Bite Registry data. Independent variables and compliance status were obtained from the registry. Logistic regression analysis was utilized on 507 dog bite patients. Results Most of dog bite patients were age group of 46–60 years old (23.1%), male (61.3%), Chinese (49.5%), seeking treatment less than 24 h after the exposure (78.3%), category two of exposure (76.3%) and bitten on lower extremities (57.8%) by an owned dog (58.4%). Only 19.5% were non-compliance to Rabies PEV. Siamese had significantly two-timed (AOR: 2.00; 95% CI 1.06, 3.76) odd higher to become non-compliance. Being bitten during 3rd (AOR: 0.27; 95% CI 0.12, 0.59), 4th (AOR: 0.24; 95% CI 0.11, 0.52) and 5th (AOR: 0.20; 95% CI 0.09, 0.44) year of the outbreak had significantly lower odds to non-comply with Rabies PEV. Conclusion 19.5% of dog bite patients still did not comply with the Rabies PEV series. Siamese would likely to non-comply whereas bitten on the 3rd, 4th and 5th years of outbreak less tendency to non-comply. Continuous health promotion to the public in the various languages despite outbreak status are ongoing to improve the perception of risk and benefit toward compliance of Rabies PEV.


Introduction
Rabies is one of the most commonly overlooked vaccinepreventable zoonotic virus diseases. It is well known for its mortality, with over 100% of instances resulting in death once the symptom appears in a human. Both domestic and wild animals can be infected with rabies. Domestic dogs, on the other hand, are blamed for the lion's share of dog-mediated human Rabies [1]. According to a study, canine Rabies caused approximately 59,000 (95% CI, 25,000,159,200) human Rabies deaths globally each year, with the Asia and 2 Methods

Study Setting
Between July 2015 and June 2020, a retrospective cohort analysis was conducted on dog bite patients who sought treatment after exposure in any healthcare facility and were recorded in the Perlis Potential Rabid Animal Bite Registry. Perlis is bounded by Thailand's Satun and Songkhla Provinces, as well as Malaysia's state of Kedah. This state has a population of 254,700 people and covers an area of 819 kms square (km 2 ) [21]. North-West Peninsular Malaysia includes state of Perlis, Kedah, Pulau Pinang and Perak. Based on study by [22], the number of canine Rabies were reported in Perlis was 20 canine Rabies whereas Kedah (18 cases) followed Pulau Pinang (4 cases). Thus, Perlis was chosen as study location as it had the most significant number of canine Rabies cases as compared to other states in North-West Peninsular Malaysia [22].

Criteria for Sample Selection
The inclusion criteria for sample selection were that the patient had a history of dog bite exposure and had received at least one Verorab vaccination. However, if a participant reported more than one exposure during the research period, only the data corresponding to the initial exposure was included; subsequent exposure reports and Rabies PEV were excluded. Patients who resumed the vaccine series despite Perlis were not included in this research.

Sample Size Determination and Sampling Method
The sample size was calculated using Power and Sample Size Calculation Software for single and two proportions with a type I error of 5% and a type II error of 20%. The calculated sample size was 486; after 10% for the data error consideration, the required sample size was 535 [23]. However, only 507 cases fulfilled the criteria for sample selection from 587 cases of rabid potential animal bites recorded in the Perlis Rabid Potential Animal Bite Registry. Thus, no sampling was applied for this study.

Data Collection and Research Tools
This study used a surveillance database of the rabid potential animal bites which was in the form of a registry. No personal information was included in the proforma to maintain the anonymization of the data set. The data were reviewed to match the inclusion and exclusion criteria, cleaned, and explored.

Perlis Rabid Potential Animal Bite Registry
The (1) Category one of exposure is defined as touching or feeding an animal, licks on intact skin, (2) Category two exposure is described as bare skin nibbling or minor scratches or abrasions without bleeding and (3) Single or numerous transdermal bites or scratches, contamination of mucous membranes or broken skin with saliva from animal licks, and exposures due to direct contact with a dog are all classified as category three exposures [9]. In this study, the site of injury was divided into three categories: upper extremities (hands, arm, elbow, and fingers), lower extremities (leg, calf, foot, and toes), and head and body (injury on the head, face, neck, chest, back, abdomen, anogenital area). This study generated factors including age group, time elapses for treatment, years of exposure, and compliance level. The time elapsed for treatment was calculated using the date of exposure and the date of vaccination start, as documented in the registry. On July 1, 2015, Perlis had the first dog bite incidents caused by a rabid dog, and a Rabies outbreak was proclaimed on the same day. As for years of exposure, we consider the first year as 1st July 2015 to 30th June 2016, followed by the 5th year ending on 30th June 2020. Despite the fact that the Rabies PEV was only administered at the government hospital, the variable of type healthcare facilities utilized in this study was defined as the first healthcare facility visited by dog bite patients for treatment or evaluation following exposure. We divide facilities into two types: primary health clinics, which include both government and private health clinics, and hospitals, which include both government and private hospitals.
On a weekly basis, the Medical Division of the Perlis State Health Department gathered the actual date to hospital for the vaccine, including both the first and follow-up vaccination series, and recorded it into the Perlis Rapid Potential Animal Bite Registry. The date in the register is utilized to establish the dependent variable status: compliance or non-compliance with the Rabies PEV. Non-compliance is defined as the dog bite patients who discontinued being vaccinated at any stage throughout the prescribed regimen (except those who discontinued vaccination after three doses, where the dog remains healthy and alive for at least ten days after the exposure). Otherwise, compliance is defined as the dog bites patients who completed the recommended course, including those who got the vaccine beyond the scheduled date of vaccination.

Statistical Analysis
The data were analyzed using SPSS version 26.0 software. Descriptive statistics were used to summarize the sociodemographic characteristics of subjects. Numerical data were presented as median (IQR) based on their normality distribution. The normality distribution was determined based on the finding of the histogram, Shapiro-Wilk test, skewness, and kurtosis. Frequency (percentage) was used to present the categorical data. The proportion of non-compliance on Rabies PEV was expressed in percentage. The nominator was the number of dog bites patients who are classified as non-compliance. The denominator was the number of dog bite patients registered in the Perlis Potential Rabid Animal Bite Registry between 1 July 2015 and 30 June 2020, whose fulfilled inclusion and exclusion criteria.
Multiple logistic regression was used for the analysis to determine the determinants of non-compliance. Simple logistic regression was performed first to select the preliminary variables that have an association with noncompliance and presented as Crude Odd Ratio (OR). Variables with p values less than 0.25 were selected for multiple logistic regression and the Backward Likelihood Ratio (LR) method was applied for this purpose. Collinearity and interaction between the significant variables as well as the Goodness of fit model were checked. The fitness of the model was tested using the Hosmer-Lemeshow goodness of fit test (p value > 0.05), the classification table (> 80%), and the receiver operating characteristic (ROC) curve (> 70%). The significance level was set at 0.05.

Results
From July 2015, when the Canine Rabies outbreak began, through June 2020, there were 587 reports of individuals bitten by rabid potential animals to the Kangar District Health Office. However, 41 records (7.0%) of people bitten by animals other than dogs, 26 records (4.4%) of people receiving no Rabies PEV at all, seven records (1.2%) of people known to continue their vaccination in a location other than Perlis, and six records (1.0%) of people experiencing 'subsequent exposure' were excluded from the study out of 587 patients. A maximum of two recorded exposures were reported for an individual. Finally, 507 patients (86.4%) were enrolled in this study.

Sociodemographic Characteristic of Dog Bite Patients
The

Proportion of Non-compliance for Rabies PEV
The proportion of non-compliance for Rabies PEV in Perlis from July 2015 till June 2020 was 19.5% (95% CI 16.1, 23.0) consisting of 99 patients. The remaining 408 patients (80.5%) complied with the Rabies PEV series for the same period in Perlis.  Tables 2 and 3 illustrate the complete findings for the simple and multiple logistic regression in this study.

Discussion
Regardless of the risk exposure of the suspected rabid animal bite especially dogs, the uttermost prevention strategy of human Rabies is early and proper administration of the Rabies PEV [9]. As in Malaysia, there were established guidelines by the Ministry of Health for appropriate management of Rabies PEV since 2015 whereby the Rabies PEV comprises 4 series of injections over 2 weeks [10]. However, there is still a barrier for the dog bite patients to complete the series. This study found that between 2015 and 2020, children under the age of 15 made up just 22.9% of all dog bite cases in Perlis. This conclusion contradicted recent research that found that children and teenagers accounted for more than two-thirds of dog bites [24,25]. Children are more vulnerable to dog bites due to their smaller size and provocative attitudes toward the dog. According to the Asian Rabies expert, this age group is less likely to disclose dog exposure to their parents, such as licking or slight scratching resulting in lower health-seeking behavior at the health care institution [26].
It is predicted that the majority of the samples will be exposed to category II, which is similar to previous research [16,23,27]. The recommended for the initiation of Rabies PEV is reserved for patients who have experienced category II or III exposure to a rabid potential dog [9,10]. Surprisingly, 30 dog bite patients (5.9%) in the current study were classed as category I of exposure yet received the Rabies PEV. Two-thirds had finished the vaccination regimen. It is difficult to distinguish the cause for the Rabies PEV commencement, whether the patient requests the vaccine, or the practitioner prescribes the vaccination despite the lower risk of exposure. Similar outcomes were seen in research on physician evaluation and PEP prescription practise [28]. As a result, practitioners emphasizing the necessity of clinical exposure evaluation for dog bite patients results in effective exposure classification and management.
The COVID-19 pandemic has resulted in a significant loss of human life throughout the world, and it poses an unprecedented threat to public health, healthcare, and the workplace [29]. Despite that, the Rabies PEV was not affected by the pandemic COVID-19 as it is an equally essential service that should be provided in the government healthcare facility. Based on the descriptive data during the 5th year of the outbreak, there was a slight reduction in the number of dog bite patients and a slight increment in the percentage of non-compliance. This reduction could be due to a lack of outdoor activity during the lockdown period which exposes the public to a dog bite or fear of attending hospital due to the risk of contracting COVID-19 after getting bitten by the dog [30,31]. The current study's 19.5% noncompliance proportion was much lower when compared to past studies in Senegal (45.5%) and India (46.8%); however, when compared to Cambodia (7.8%), this study's findings were higher [15,17,27]. In Senegal, the cost of a complete Rabies PEV course without RIG was around 60 Euro, or 37.5% of the average monthly disposable wage (160 Euro) based on the World Development Indicator for Senegal [15]. Rabies PEV is offered free of charge in government hospitals in Malaysia. Aside from that, in Cambodia, the Rabies PEV is subsidized rather than free. As a result, because they have already paid the entire fee upfront, it encourages the patients to complete the vaccination series [27].
There is a discrepancy in the Rabies PEV regime between India and Malaysia. They employed the Updated Thai Red Cross intradermal post-exposure rabies prophylaxis regimen, which is lengthier (across 28 days) and so had a greater risk of noncompliance [17]. In Malaysia, the standard protocol for administering Rabies PEV consists of four injections spaced out over 14 days [10]. Shortening the entire course of Rabies PEV's length will assist improve patients' compliance, minimizing the burden on patients in terms of lost work time and transportation expenditures [26,32]. This hypothesis explains why Malaysia has a lower proportion of non-compliance as compared to India.
This study also shown that ethnicity was a determinant in noncompliance with Rabies PEV. When compared to other ethnicities, Siamese people are twice as likely to be non-compliant. According to the Malaysian Population and Housing Census of 2000, an estimated 60,000 Siamese ethnic people accounted for less than 1% of Malaysia's population [33]. This minority ethic may struggle to understand Rabies' health education materials, which are only available in Malays or English, resulting in poor health literacy. Previous research in North Carolina found evidence of low Rabies awareness among minorities as a result of similar issues [34]. Several of them hold Thai citizenship are also married to Malaysians. Previous research revealed that cultural beliefs impacted their perspective of health and treatment choices [35]. As a result, we anticipate that non-Malaysian citizenship Siamese ethnic may return to their homeland after having first treatment in Malaysia, resulting in non-compliance with the Rabies PEV series. Furthermore, our study showed that years of exposure are associated to noncompliance with Rabies PEV. However, it is not one of the determinants that affect noncompliance with the Rabies PEV. When adjusted for ethnicity, dog bite patients bitten from 1st July 2017 to 30th June June 2018 (third year of the Perlis Rabies outbreak) were 73% less likely to become non-compliance than those bitten from 1st July 2015 to 30th June 2016, which was the first year since the Perlis Rabies outbreak was declared. Finally, the odds of receiving an incomplete Rabies PEV series were lower from 1st July 2018 to 30th June 2019, the fourth year of the Perlis Rabies outbreak (by 76%), and from 1st July 2019 to 30th June 2020, the fifth year of the Perlis Rabies outbreak (by 80%). Thus, our findings contradicted recent research conducted in Bhutan and Cambodia, which found that those bitten later in the study period were more likely to have an incomplete series of the Rabies PEV [13,27].
The findings in this study might be explained by the fact that human Rabies cases and outbreaks in Sarawak, Malaysia were reported on July 1st, 2017. During the third year of the Perlis Rabies outbreak, 11 human Rabies cases were documented in Sarawak [36,37]. Following that, eight and five human Rabies cases were reported in Sarawak's Rabies outbreak during the fourth and fifth years of Perlis' Rabies outbreak [38,39]. Sarawak is located on the island of Borneo, which is not part of Peninsular Malaysia. Malaysia (Sabah and Sarawak), Brunei, Kalimantan, and Indonesia make up the Borneo islands. Although Sarawak is not part of peninsular Malaysia, human Rabies is a national health concern that has received extensive coverage in the media and on social media. Over the last three years of data collection, the Director-General of Health Malaysia's regular updates through mass and social media served as a wake-up call to Perlis residents about the necessity of Rabies PEV's completion. As a result, it is likely explanation for why compliance improved in years 3, 4, and 5. Messages in the media are effective at raising awareness and increasing demand for the Rabies PEV [40,41]. This mass media reportage, when applied in combination with the health belief model, enhances the perceived of vulnerability and severity of dog bite patients, resulting in positive changes in Rabies PEV compliance.

Study Strength and Limitation
This study sought to be a comprehensive study that incorporates data provided by various institutions within the Perlis State Health Department on handling dog bite cases, which contributes the most to this study's strength.
Data for the registry were acquired through primary care (government and private), hospitals, and field investigations.
As a consequence, our study outperformed a previous study that was confined to a single institution [12,15,42]. Our findings are beneficial not only to the health department but also to other relevant authorities, especially those in the one health approach. For example, the Veterinary Service Department used our localized demographic of dog bite patients to strategize the target population priority for the dog bite preventions besides utilizing these findings to encourage the dog owners to vaccinate their pets. The limitation of our study on the determinants of noncompliance with Rabies PEV is that we did not conduct an in-depth interview with patients to determine why they did not comply with the Rabies PEV as indicated. This compliance behavior is generally the result of complex social, economic, cultural, accessibility, and other interacting elements that a basic categorical descriptive database cannot reflect on, making it impossible to reflect on the causes of non-compliance with the Rabies PEV. This approach supplements the qualitative or mixed-method study design to provide a more comprehensive picture of the local context on noncompliance behavior, which is recommended in the future. Because this investigation was intended as a retrospective cohort study utilizing secondary data, there is a dearth of information in the registry on certain variables such as education status, citizenship, socioeconomic level, and type of exposure. To establish evidence on the risk factors, it is advised that this study be conducted in a prospective cohort. Ideally, this would allow healthcare practitioners to identify dog bite patients at the time of reporting who fulfill the determinants that place them at a higher risk of not completing the course, allowing for targeted assistance. Nonetheless, our research has persuaded policymakers that, even with services supplied at practically no cost, one-fifth of our population may take the Rabies PEV series for granted. Our findings on the determinants of noncompliance on Rabies PEV may be used as a slice of scientific evidence for future promotion planning to attain elimination status by 2030.

Conclusions
Despite the vaccine being available, accessible, and affordable most of the time in Malaysia, a segment (19.5%) of the population in North-West Peninsular Malaysia did not comply with the Rabies PEV series, putting them at a higher risk of developing human Rabies. Dog bite patients of Siamese ethnicity are more likely to possess an incomplete Rabies PEV series, although those bitten in the third, fourth, and fifth years of the outbreak are less likely to become noncompliant. As a result, it is advised that health education materials be prepared in a variety of languages, including Thai. Continuous public advocacy of Rabies prevention by top Ministry of Health officials, regardless of outbreak status, is needed to boost the perception of risk and benefits toward Rabies PEV compliance.
Ethics Committee Universiti Sains Malaysia with the research code of SM/JEPeM/20110590 on the 17th December 2020.
Patient Consent for Publication Not applicable. All records were retrieved from the Perlis Rabid Potential Animal Bite Registry were fully anonymised before researchers accessed them. However, the permission to utilize this data were obtain from the Director General of Malaysia and Director of Perlis State Health Department as the data custodian.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third-party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.