Predictors of Health Care Seeking Pace Among Newly-Diagnosed Pulmonary Tuberculosis Patients In Cebu City, Philippines: A Cross-Sectional Study

Tuberculosis remains a major public health dilemma in the Philippines. While free and effective TB diagnosis and treatment have been made available since 1996, a number of patients still delay accessing them re sulting to increased TB mortality and community transmission. This study is aimed at determining the significant variables that can predict a TB patient’s pace in seeking appropriate health care. Methods A descriptive, correlational cross-sectional survey was done to 127 newly-diagnosed TB patients from 10 government-owned TB DOTS facilities in Cebu City. Participants were categorically grouped into Prompt Health Seekers and Delayed Health Seekers using a norm-referenced median dichotomy. Descriptive and inferential analysis utilized included Chi-square Test, Pearson’s Correlation and Discriminant Analysis.


Abstract Background
Tuberculosis remains a major public health dilemma in the Philippines. While free and effective TB diagnosis and treatment have been made available since 1996, a number of patients still delay accessing them re sulting to increased TB mortality and community transmission. This study is aimed at determining the significant variables that can predict a TB patient's pace in seeking appropriate health care.

Methods
A descriptive, correlational cross-sectional survey was done to 127 newly-diagnosed TB patients from 10 government-owned TB DOTS facilities in Cebu City. Participants were categorically grouped into Prompt Health Seekers and Delayed Health Seekers using a norm-referenced median dichotomy.
Descriptive and inferential analysis utilized included Chi-square Test, Pearson's Correlation and Discriminant Analysis.

Results
A mean health care seeking delay of 59 days (median=49) was recorded. Most of the participants experienced unexplained cough which they perceived as only slightly dangerous. 38% of them reported not knowing any symptom related to TB while 39% were not aware of the free public TB services. Most patients have made 2-3 prior health recourses, mostly through self-medication, before finally contacting a DOTS facility. Marital status (p=0.004), the number of symptoms experienced (p=0.000), first symptom experienced (p=0.016), perceived dangerousness of all symptoms experienced (p=0.009), perceived dangerousness of the first symptom experienced (p=0.001), perceived social stigma (p=0.035) and, perceived social support (p=0.002) were found to be significantly associated with the patients' health care seeking pace. Extending to multivariate analysis, five independent variables namely marital status (p=0.037), number of symptoms experienced (p=0.018), perceived dangerousness of all the symptoms experienced (p=0.028), perceived social stigma (p=0.026) and, perceived social support (p=0.001) can significantly predict the patients' health care seeking pace.

Conclusion
A considerable health care seeking delay of 59 days was documented, especially among those with lower perception of social support, higher perception of social stigma and those who do not currently have partners. Patients who experienced more symptoms and those who view these symptoms as less dangerous were, however, likely to seek immediate health care. Reducing health care seeking delays through intensified health information campaigns, strengthening social support systems and reducing social stigma is recommended. Background Tuberculosis (TB), a bacterial infectious disease, has become a major public health dilemma which prompted the World Health Organization (WHO) in 1993 to pronounce it as a worldwide public health crisis. Nineteen years after, tuberculosis still remains an enormous global burden with 8.6 million new TB cases and 1.3 million TB-related mortalities [1]. Following the human immunodeficiency virus, the disease ranked second as the primary cause of mortality related to an infectious disease globally [1].
Locally, the Philippine TB scenario is parallel to the global pandemic with an incidence rate as high as 322/100,000 population [1]. In 2010, TB holds the sixth spot in the ten leading causes of death in the country accounting to 5.1% of the total deaths [2]. WHO ranked the Philippines ninth among the countries with high TB prevalence worldwide and the highest in Southeast Asia. An estimated 324,000 Filipinos living with tuberculosis, resulting to more than 14,440 recorded mortalities in 2015 [3].
To address the growing concern, the Philippine Department of Health (DOH) through the National Tuberculosis Control Program (NTP), has employed the WHO-endorsed Directly Observed Treatment Shortcourse (DOTS) Strategy in 1996. While the DOH recorded impressive treatment success rates reaching national targets, case detection rate for all forms of TB is only at 82% below the 90% national target. Case detection rate by region also varies from 46% to 114% [2]. This means that although the DOTS strategy has been effective in treating TB cases, there are still a number of missed TB patients who were either delayed in diagnosis or not diagnosed at all.
Delays in the diagnosis and treatment of TB cases are major impeding factors in TB control.
Diagnostic deferral of pulmonary tuberculosis can accelerate the progression to a more extensive disease [4], increase the risk of complications and mortality [5] and contribute to the increased transmission of tuberculosis in the community [6]. WHO affirms that around ten to fifteen individuals will be infected by a person who has untreated tuberculosis [1].
While there has been no agreed definition of what can be considered an acceptable length of health care seeking delay in the Philippines, DOH has proposed that cough for two weeks, without underlying risk factors is the appropriate time to seek health care [2]. Literature review on studies done on diagnostic delay from 1995-2004 also pointed out that a 2-3 week delay is acceptable as patient delay in the TB control point of view [7].
A plethora of key patient factors and health system-related variables have been attributed to the delays in TB diagnosis and treatment. Factors such as patient's socioeconomic status, distance between the patient's residence and the health service provider, social stigma, level of disease awareness, personal perception of the disease, severity and seriousness of the disease, and health personnel attributes are just a few [8]. Identifying, analyzing and addressing the numerous factors that are contributory to patient and system delay are essential in order to devise and formulate strategies for all partners involved in tuberculosis control [8].
With these considerations, this study is embarked to assess the length of health care seeking pace among newly-diagnosed TB patients in Cebu City and identify which among the 26 variables clustered into 7 groups: 1) Socio-demographic Profile (age, gender, educational attainment, marital status, employment status, income, history of smoking), 2) Symptom-related variables (symptoms experienced, perceived dangerousness of symptoms experienced), 3) Accessibility to TB DOTS Facility (distance to TB DOTS facility, travel time to DOTS facility), 4) Cognition-specific Variables (knowledge on TB symptoms, knowledge on someone with TB, awareness of available TB services), 5) Prior Health Recourses, 6) Intrapersonal Variables, 7) Social Influence has a significant relationship and can therefore predict their health care seeking pace (HCSP).

Study Design
The study employed a descriptive correlational cross-sectional survey design.

Study Sites
The study was conducted in ten (10) DOH-recognized and NTP-affiliated public TB DOTS facilities in Cebu City, a highly-ubanized city located in Central Philippines. Site selection was done through stratified random sampling by randomly choosing two facilities from each of the five health areas. Complete enumeration of all TB patients enrolled to DOTS treatment in the 10 facilities from March to June 2017 was then be employed to determine the study's final respondents and sample size.

Study Population
127 eligible research respondents participated based on the criteria that they were newlydiagnosed bacteriologically-confirmed or clinically-diagnosed pulmonary TB cases 15 years old and above. Newly-diagnosed TB patients are those "who have no previous history of anti-TB medication intake or have taken TB medications for less than one month". The study excluded patients less than 15 years old since in the Philippine context, they are referred to as TB in children who follow a totally different diagnostic algorithm and protocol in comparison to adults aged 15 years old and above [2].
Patients who were accidentally diagnosed to have tuberculosis, those who did not present any symptoms during contact with the TB DOTS facility, extrapulmonary TB patients, and those who have had history of treatment for more than 2 months were also excluded. In addition, only patients currently enrolled to TB DOTS treatment for less than 56 days were used in the study to ensure fresh recall of the details surrounding the health care seeking encounter and to minimize recall bias.

Research Instrument
The researcher utilized a combination of a researcher-made and a standardized questionnaire. The tool was translated to the local Cebuano dialect for easy comprehension. Pretesting and back translation was also done to ensure validity and reliability. Comments from the pretesting respondents were collected and incorporated to the instrument as deemed appropriate.

Study Limitations
The present study have limitations. First, the data gathering procedure used in measuring HCSP was through self-reporting which implies a recall bias. To minimize this problem, the study specifically asked first about the onset of the first symptoms (most commonly cough) and the specific date of their first TB DOTS facility consultation/ encounter. The researcher also used a local calendar listing the main national celebrations and holidays vis-à-vis the dates when the sputum collection and TB treatment was done as indicated in the TB Registry, in order to clearly refresh the patient's memory of the date of appearance of the first symptoms. Second, the measurement of distance from the patients' residence to the TB DOTS facility is merely through estimation due to lack of appropriate technology. Third, because of budgetary constraints, the number of sample facilities were limited to only two per health area.

Data Analysis
The study participants were categorically grouped into Prompt Health Seekers and Delayed Health Seekers. Since there is no agreed definition of what can be considered an acceptable length of health care seeking delay in the Philippines, the generated median value was used to dichotomize the participants into two classifications, analogous to a previous study [9]. Using the norm-referenced median dichotomy to classify the participants, the statistical treatments included both descriptive and inferential. The Microsoft Excel was used to determine frequency counts, percentages, central

Socio-demographic Profile
As reflected in Table 1, of the one hundred twenty-seven (127) TB patients who participated in the study, majority were males (71%), aged 19-40 years old (46%), have had at least secondary education (74%) with 25% of them graduating high school, has a spouse/partner (64%) and are employed (52%) (Complete set of raw data stored in Additional File 1). About 46% of the respondents fall under the fourth income class quintile receiving a monthly income within the range of Php 8,333.26 to Php 20,833.25. On average, the monthly income of the patients is pegged at Php 7,896.83 which is way below the poverty threshold of Php 9,064 [10]. In regards to smoking history, 58% of the respondents reportedly quitted smoking.
However, only two of the five top presenting symptoms were respiratory-specific symptoms (cough and chest pain). On their perception as to the dangerousness of the symptoms they experienced, patients regard them as slightly dangerous with a mean of 2.44. Results further show that although unexplained cough, weight loss, fatigue/malaise and chest/back pain were the top presenting symptoms, the respondents only view them as slightly dangerous. Only bloody sputum (hemoptysis) and pulmonary TB-consistent chest radiographic result were perceived as very dangerous while shortness of breath/difficulty of breathing (dyspnea) as moderately dangerous. Grouping the respondents according to the number of symptoms experienced, more than half of the patients belong to the group with a few (1-3) symptoms. Patients with seven or more symptoms only account for 10% of the total respondents. On average, patients experienced 4 symptoms. Unexplained cough also placed on the top spot as the first symptom prompting health care seeking. It is interesting that unexplained cough ranked on the top spot of the symptoms that prompted health care albeit the fact that patients view the symptom as slightly dangerous.

Accessibility to DOTS services
The average reported distance from the respondents' residence to the TB DOTS facility is 736 meters. Stratifying them to three categories, 50% of the respondents live within the 500 meter radius from the TB DOTS facility. The average travel time to the facility is 13 minutes with majority of the respondents (88%) living less than 30 minutes away.

Cognition-related characteristics
38% of the repondents reported not knowing any TB-related symptoms, while 42% of them revealed knowing unexplained cough as a hallmark symptom of tuberculosis. Two (2) symptoms is the average number of TB-related symptoms known by the respondents. 57% of them do not know someone who has TB while 39% of them also declared that they were not aware of the availability of the public TB services prior to their first contact with the TB DOTS facility.

Prior Health Recourses
Only 2.4% of the respondents have not resorted to any form of health recourse prior to contact with the DOTS facility. 51% of them resorted to self-medication of unprescribed, over-thecounter drugs; 25% took market-available food supplements, and; 24% visited traditional health providers such as herbolaryos, hilots and faith healers. Although the data does not represent a majority, the huge proportion of patients patronizing non-medical health options is not negligible.
The number of prior health recourses made ranged from 1 to 6 health recourses with an average of 2.65~3. Most of the patients made an average of 1.34 medically-related health recourses compared to less than one non-medical health recourse (Mean 0.75). 31% of those who engaged in medically-related recourses, opted to directly visit a private medical clinic with general practitioner as attending physician.
Consequently, the study also analyzed which health facilities were accessed first before the other, based on 1) the type of facility (clinic vs government-recognized health facility) and 2) the ownership of the facility (public or private). 46% of the patients went first to government-recognized health facilities compared to the 35% who went first to clinics. In terms of ownership of health facility, 44% of the patients went to a public health facility/clinic first in contrast to the 38% who accessed a private health facility/clinic first.

Intrapersonal Variables
With a weighted mean score of 3.55, respondensts view the services of the facility as Very Effective. The respondents also have a low perception of social stigma with a weighted mean of 2.86.
Dissecting into the four subcomponents of social stigma perception, 1) alienation (x̅ = 3.14), 2) stereotype endorsement (x̅ = 2.73) and 3) perceived discrimination & social withdrawal (x̅ = 2.78) are classifiable as slightly low based on their weighted mean. Stigma resistance, the fourth component, recorded a non-reverse-coded mean of 4.04 classifiable as slightly high. The respondents also have a slightly high perception of social support with a weighted mean score registering at 4.18. Looking into the subcomponents of social support, 1) appraisal support ((x̅ = 4.32), 2) belonging support ((x̅ = 4.07), and 3) tangible support (x̅ = 4.16) are all classifiable as slightly high.

Social Influence
74% of the patients reported that they were influenced by their immediate family members to seek care to the TB DOTS Facility. Local health staff and media were also identified by 10% and 5% of the patients, respectively, as their primary driver for seeking health care.

Respondents' health care seeking pace (HCSP) and their classification
The mean health care seeking pace (HCSP) is 59.05 days. The shortest is 1 day while the longest is more than a year. The generated median HCSP is 49 days. The median HCSP, utilized as the reference point, dichotomized the respondents into two categories, 1) Prompt Health Seeker whose HCSP range is 1-49 days, and 2) Delayed Health Seeker HCSP range is 50-367 days. 51% of the respondents were classified as prompt health seekers while 49% of them delayed health seekers. perceived social support (p=0.002) were also significantly associated with the respondents' HCSP.

Significant Predictors of Health Care Seeking Pace
Although the univariate correlation analysis yielded significant relationships between the seven (7) variables, when extended to multivariate discriminant analysis, only five (5)   repercussions to the patient as well as the community as these delays can hasten the progression to more complex illness [4], escalate the risk of complications and mortality [5] and amplify TB transmission to the community [6]. The generated median HCSP is also long at 49 days, similar to the results found in Hongkong [13]. However, this is shorter than in Ethiopia, 43 weeks [6] and France 68 days [14], and longer than in Zimbabwe 36 days [15] and China 6 days [9]. This delay in health seeking might be due to a number of factors ranging from socioeconomic status, distance to health services, social stigma, personal perception of the disease and severity and seriousness of the disease [8].
In terms of gender, the proportion of men afflicted by the disease is higher compared to women. Similarly, DOH noted that while both sexes are at risk of contracting tuberculosis, the infectious illness is more prevalent among males [2]. Reasons for this imbalance might include greater prevalence of TB associated with men or it may indicate a persistence of personal barriers, societal inequities and health system challenges disproportionately influencing timely TB diagnosis among women [8]. Less specific clinical TB presentations are also more common among women i.e.
fewer characteristic symptoms, such as blood in sputum [8] which may cause such gender disparities.
However, gender is not found to be a significant predictor of HCSP in contrast to a study in Nigeria [16]. Income, tagged in many studies [12,17,18] as a good predictor of patient delay, was not a significant determinant in this study. While the majority of the respondents (63%) are living below the poverty threshold whose income falls below Php 9,064 [10], it did not seem to have an effect on their health care seeking. Furthermore, the results also negate the findings of studies stating that HCSP is dependent on the influence of other patient socio-demographic variables such as age [13,16,19,32], educational attainment [12,18,20] and employment [13,21] and history of smoking [22]. It can be implied that health care seeking goes beyond sociodemographic elements and that other factors may have a more powerful influence over a decision to seek health care.
Accesibility variables namely 1) distance to DOTS facility (p=0.553) and, 2) travel time to DOTS facility (p=0.083) have also no significant relationship to the patients' HCSP. Patients living near or far from the TB DOTS facility may have equal risk of delaying their HCSP, a finding that is different to some published studies [16,23,24,25] Cognition specific variables such as 1) knowledge on TB symptoms (p=0.686); 2) knowledge of someone with TB (p=0.507), and; 3) Awareness of available TB services (p=0.309) were also not significantly associated with HCSP. Although several literatures identified knowledge of TB symptoms [18,21,26] and awareness of available TB services [5,27,28] as fundamental variables with a huge impact on how patients seek health care, it is not the case in Cebu City. It is alarming however to note that only 2 tuberculosis-related symptoms were known by the respondents. 39% of them also declared not being aware of available free public TB services prior to their first contact with the TB DOTS facility. Public dissemination of correct TB information and the availability of TB DOTS services might either be too low or that the efficiency of information penetration to these intended population was poor. Paradoxically, TB DOTS facilities were reported to be located less than 1 kilometer and less than 30 minutes from the patients' residence, an advantage that could have been used to intensify health education and information dissemination.
Prior health recourses variables such as 1) type of prior health recourse mostly utilized (p=0.059), 2) type of first prior health recourse (p=0.231), 3) number of prior health recourses made (p=0.321), 4) type of first health facility visited (p=0.326), and 5) ownership of first health facility visited (p=0.685) have no significant relationship to a patients' HCSP. Therefore, prompt and delayed health care seeking may be observed in patients regardless of whether they made prior recourses on their symptoms, negating the findings of previous studies [6,8,15,16,18,23]. Although variables under prior health recourses did not seem to influence HCSP, the huge proportion of patients patronizing non-medical health options is alarming. The high utilization rates for non-medical choices may suggest that there could be 1) poor information dissemination on the available medical services; 2) presence of deeply-founded cultural & traditional beliefs, 3) poor awareness on the effect of nonmedically suggested alternatives, and; 4) presence of economic barriers that hinder patients from accessing appropriate scientific treatment. Multiple health recourses were also made prior to contact with the TB DOTS facility, placing an impact to the total length of health seeking and diagnostic delays, as revealed in many studies [8,11,21,29]. Most of the first health recourse made is still selfmedication, implying a widely-accepted culture of self-medication among Cebuanos, a relatively easy access to pharmaceutical drugs or a pervasive poor knowledge on ill-effects of self-medication. Selfmedication for TB-related symptoms is highly correlated with patient delay [6,15,18] [32,33]. The respondents' partners probably provided an additional push, be it in the form of physical, psychological or emotional support, for them to seek consultation and treatment.
Four symptom-related variables showed a significant relationship to the patient's HCSP, namely, the 1) number of symptoms experienced (p= 0.000); 2) first symptom experienced (p=0.016); 3) perceived dangerousness of all the symptoms experienced (p=0.009), and; 4) perceived dangerousness of first symptoms experienced (p=0.001). Analysis shows that as the number of symptoms increases, the likelihood that the patient delays seeking health care also increases. Therefore, the number of symptoms has a deterring and delaying power over HCSP. Some patients wait for the appearance of more symptoms before they are convinced to have themselves checked. Waiting for the appearance of many symptoms causes further delays in the diagnosis and treatment. Another perspective that might be considered in this analysis is to look at the HCSP as the variable that affects the appearance of the number of symptoms. As the patient, regardless of the reason, delays diagnosis and treatment, the number of symptoms he experiences also increases. This corroborates with the findings of many studies that delaying diagnosis and treatment may pose additional symptomatic burden to the patients, accelerate risk towards developing a more advanced disease [4,34] and increase the possibilities of complications and mortality [5,34].
Another finding showed that most of the respondents who had 1) unexplained fever (100%), 2) shortness of breath (83%), 3) chest/ back pain (83%) and pulmonary tuberculosis-consistent chest X-Ray (75%) as their first symptom were classified as prompt health seekers. Most of the respondents who had fatigue/ malaise (100%) as the first symptom, on the other hand, were classified as delayed health seekers. The type of symptom first experienced by the patient might influence his HCSP.
Shortness of breath, chest/ back pain and pulmonary tuberculosis-consistent chest X-Ray are respiratory specific symptoms while unexplained fever and fatigue or malaise are general symptoms.
The results somehow affirm the results of previous studies that presence of less specific symptoms are associated with longer delays (13; 35].
The increased perception in the dangerousness of all the symptoms experienced (p=0.007) also increases the likelihood of delay as did the increase in the perceived dangerousness of the first symptom (p=0.001). This implies that increased perception of dangerousness is a deterrent to prompt health care seeking, a finding that runs in contrast to other studies [28]. Some patients equate the perception of danger in the symptoms they experience with intense fear of having a severe disease, a possibility of prolonged hospitalization or risk of incurring additional financial expenses. These discouraged them from accessing appropriate health care, as evidenced by an anecdotal statement made by one patient: "Mahadlok ko sa akong sintomas pero mas mahadlok ko mupakonsulta kay basin og nay makit-an nga grabe nga sakit. Makagasto pa unya ug dako ang akong pamilya". internal stigma translates to longer HCSP. The finding suggests that high perception of social stigma among TB patients prevents them from seeking timely medical care. The findings further verify results of many studies (8,12,21]. Health seeking is hampered by the presence of social stigma because it can inflict damage to the person's social status, social relationships and even income source [36]. Analysis of the data also revealed that perceived social support (r-value=-0.278) has a weakly negative correlation with HCSP, implying that an increase in the level of the patient's perceived social support decreased the tendency to delay HCSP. Patients with high levels of social support are highly likely to engage in prompt health care seeking behaviors [37]. As reflected in Table   1, perceived social support in Cebu is high implying that patients' circles such as partners, family, friends and the community might have been explicitly supportive making it tangible to the patient. In the Philippine society, the Filipino virtue ethics is richly "relationship-oriented" which is established on the two basic foundations of the country's culture, "loob" which is interpreted as "relational will" and "kapwa" which is better understood as "together with the person" [38], which may have a relative impact on the patients' perceived assessment of social support. The high perception of social support may also explain the relatively low perception of stigma among the respondents.
Extending to multivariate analysis using discriminant analysis, only five (5) independent variable (marital status, number of symptoms experienced, perceived dangerousness of all the symptoms experienced, perceived social stigma and perceived social support) came out as the predictors of HCSP. The results further strengthen the concept that a supportive, non-stigmatizing and accepting social environment, is fundamental to promoting positive health seeking behaviors among presumptive of TB patients. The presence of a physically tangible source of support such as a spouse or a partner further reinforces this health promoting social environment. This confirms the findings in literatures where explicit social support from social relationships and social networks serve as a potent starting point and driver for a causal flow of positive health outcomes [39]. The relationship between the number of symptoms and HCSP implies that patients do wait for the appearance of more symptoms before making contact with TB DOTS facilities. A higher perception of danger over the symptom experienced also derails further action which may be associated with intrapersonal and psychological impacts of apprehension or fear to health care seeking

Conclusion
There is a considerable delay (59 days) in the HCSP among TB patients in Cebu City using both the norm-referenced median dichotomy (49 days) and the DOH advocated timeline of 14 days. Patients who delay seeking health care are those with lower perception of social support, higher perception of social stigma and those who do not currently have partners (single, widowed/widower, divorced/separated/annuled). Moreover, patients who have experienced more symptoms and those who view these symptoms as less dangerous are patients who would readily seek health care.
Reducing health care seeking delays through intensified health information campaigns, strengthening social support systems and reducing social stigma is highly recommended. respondents were asked to sign an informed consent written in both English and local Cebuano dialect, indicating their agreement to participate in the study. On the other hand, legal Parents/ Guardians of respondents below the age of 18 years old were required to co-sign an assent form together with the respondent, in compliance with Philippine national laws. It was be emphasized during this phase that the study results will be kept confidential and their anonymity will be of prime consideration. The respondents were also given liberty to withdraw at any point of the study.

Consent for publication
Not applicable

Availability of data and materials
All data generated or analysed during this study are included in this published article and its supplementary information files.

Competing Interests
The author declares that he has no competing interests.

Funding
The funding source of this study is purely derived from the author's personal finances.

Author's contribution
The author is the sole investigator of this study from conception, analysis, interpretation and the development of this manuscript.