F. Maamari1
كشف حالات السل: التأخَّر في التشخيص والمعالجة ومحدداته
فاديا معماري
الخلاصـة: قيَّمت هذه الدراسة التي أجريت في الجمهورية العربية السورية، تَوَاتُرَ ومحدِّدات التأخر في تشخيص حالات السل الجديدة الإيجابية اللطاخة، في مراكز المعالجة القصيرة الأمد تحت الإشراف المباشر DOTS. وبيَّنت هذه الدراسة التي شملت 800 مريض، أن متوسط التأخر المتعلِّق بالْتِمَاس الرعاية قد بلغ 52.7 يوماً (المجال: من 15 إلى 698)، ومتوسط التأخر المتعلق بإجراءات النظامٍ الصحي السابقة للتشخيص قد بلغ 24.8 يوماً، جعل متوسط مدة التأخر الكلِّي قبل التشخيص 77.6 يوماً. في حين كان متوسط الفترة بين التشخيص وبين بدء المعالجة قصيراً جداً إذ لم يتجاوز 2.9 يوماً. وتمثَّلت عوامل الاختطار الهامة المسؤولة عن التأخر الكلي في ما يلي: العيش في أماكن بعيدة عن المرفق الصحي، والشعور البالغ بالوصمة، والْتِمَاس الرعاية في البداية لدى شخص من غير مقدِّمي الرعاية الصحية، ومراجعة أكثر من مرفق صحي واحد قبل التشخيص.
ABSTRACT: This study in the Syrian Arab Republic assessed the frequency and determinants of delays in diagnosis and treatment of new smear-positive tuberculosis cases at DOTS treatment centres. Among 800 patients, the mean delay due to patient care-seeking behaviour was 52.7 days (range 15–698) and the health system delay before diagnosis was 24.8 days; thus the mean total delay before diagnosis was 77.6 days. The mean delay from diagnosis to start of treatment was very short at 2.9 days. Significant risk factors for total delay were: living far from the health facility, feeling a high degree of stigma, seeking initial care at a non-health care provider and having more than 1 health care encounter before diagnosis.
Dépistage de la tuberculose : les retards au diagnostic et au traitement et leurs déterminants
RÉSUMÉ: Cette étude, conduite en République arabe syrienne, avait pour objectif d’évaluer la fréquence et les facteurs déterminants des retards dans le diagnostic et le traitement de nouveaux cas de tuberculose à bacilloscopie positive (TPM+) dans les centres DOTS (traitement de brève durée sous surveillance directe). Sur un effectif de 800 patients, le retard moyen dû au comportement de recours aux soins du patient était de 52,7 jours (fourchette : 15-698) et le retard du diagnostic imputable au système de santé était de 24,8 jours, soit un délai total moyen avant diagnostic de 77,6 jours. Le délai moyen entre l’établissement du diagnostic et l’instauration du traitement est apparu très bref, à savoir 2,9 jours. Si l’on considère le retard global, les facteurs de risque significatifs sont l’éloignement du domicile par rapport au centre de soins, le poids de la stigmatisation, le recours dans un premier temps à une personne autre qu’un prestataire de soins et la multiplication des consultations médicales avant l’établissement du diagnostic.
1National Programme of Tuberculosis, Ministry of Health, Damascus, Syrian Arab Republic ( Correspondence to F. Maamari:
Received: 13/07/04; accepted: 18/07/05
EMHJ, 2008, 14(3):531-545
Introduction
In 1993, the World Health Organization (WHO) declared a state of global emergency for tuberculosis (TB), due to the steady increase of the disease worldwide [1]. In 1995, the directly observed treatment, short-course (DOTS) strategy was established as a key plan to achieve TB control worldwide. The global targets of TB control are to achieve 70% case detection and 85% cure rates by 2005. According to the latest WHO global report [1], DOTS programmes successfully treated 82% of all registered new smear-positive patients in 2002, but detected only 45% of the estimated TB cases in the world in 2003. The report indicated that the target of 70% case detection might not be reached until 2015, unless interventions were used to increase the case detection rate.
The situation in the Syrian Arab Republic is similar. In 2003, 1545 cases of new smear-positive pulmonary TB were detected, and 1353 of them were successfully treated [2]. As the estimated incidence of TB in the country is 20/100 000 population, the case detection rate is 44%, while the treatment success rate is 88%, and DOTS coverage of 100% was reached in April 2000. Therefore improving case-finding is a priority, and operational research is needed.
One important research area in this regard is the assessment of the delays between the start of symptoms and making the diagnosis of TB and between making the diagnosis and starting treatment, and what causes delays. The delay in diagnosis and treatment of TB patients can be caused by patients or by the health system, or both. The problems of delay in TB case-finding have been studied to some extent in other countries, both in developing and in developed countries, since delays in diagnosis have been noted in both high and low TB prevalence countries [3–6]. In high prevalence countries delays build up due to prolonged patient and doctor delays [3]. Several factors have been identified as influencing delay in diagnosis and start of treatment including the individual patients’ perception of the disease, the severity of the disease, patients’ access to health services and the expertise of health personnel.
Some previous studies have suggested that the determinants of a longer delay include specific patient groups (e.g. women in Viet Nam and Nepal, rural residents in Tanzania, nationality in Los Angeles, USA [3–6]). Other studies suggest that the most important factors for delays were availability and accessibility of health services. All these studies highlighted the importance of delay in increasing costs and mortality due to TB. The multiple factors causing delay in diagnosis must be clearly identified and addressed locally in order to improve the quality and effectiveness of the national TB control programmes (NTPs). Studies of case-finding, particularly analysing delays and their determinants, would allow good assessment of case-finding success under DOTS.
This study of TB case-finding in the Syrian Arab Republic made an in-depth analysis of various types of delay and their determinants. The goal was to identify gaps in case-finding under DOTS in order to assist in planning future interventions.
Methods
A cross-sectional study was conducted in all 13 NTP centres implementing DOTS in the country. The study covered new patients seen during the period 1 February 2003 to 30 September 2003. The number enrolled from each centre was proportional to the number recorded in that centre in 2001.
A total of 800 new smear-positive pulmonary TB patients aged more than 15 years old were interviewed consecutively according to a structured and pre-tested questionnaire. The sample size was based on the estimated incidence of TB. The questionnaire included information about the time intervals between onset of symptoms and each of the following events: initial health-seeking, first visit to a health care provider, making the diagnosis of TB and starting DOTS treatment. Health workers, including doctors and paramedical staff, were given intensive training on interview and probing techniques. They interviewed the patients during the first 2 weeks of their treatment, after obtaining informed consent.
The patients were also interviewed about factors that might influence health-seeking behaviour and accessibility to timely and appropriate care:
Sociodemographic status (measured using a summation score of education, occupation and incomes; best = 0 and worst = 7).
Satisfaction with care (measured on a 4-point Likert scale; 0 = best and 3 = worst). The variables included availability of services in TB centres, prompt action from primary care personnel, adequacy of equipment and free medication in these centres, proper coverage of TB centres in the area, health facility workload, and waiting time. The cut-off for satisfied/unsatisfied was the median value.
Knowledge (measured on a 3-point Likert scale; 0 = best and 2 = worst). The variables included knowledge about the type of disease, its causes, curability, existed of a vaccine, type of anti-TB drugs and duration of treatment. The cut-off for adequate/inadequate knowledge was the median value.
Feelings of stigma about TB (measured on a 5-point Likert scale; 0 = highest, 4 = lowest degree of stigma). The variables included feeling ashamed of having TB, having to hide TB diagnosis from others and having problems with family relations, work performance, marriage prospects, family responsibilities, infertility, pregnancy or breastfeeding. The cut-off for high/low stigma was the median value.
Definitions
The following definitions of delay were used:
Diagnostic delay: time interval between onset of symptoms and diagnosis of pulmonary TB. This consists of 2 components:
Patient related diagnostic delay: time interval between onset of symptoms and first seeking care at a health care provider.
Health system related diagnostic delay: time interval between seeking care at a health care provider and diagnosis of pulmonary TB.
Treatment delay: time interval between diagnosis and start of DOTS treatment.
Health system delay: time interval between seeking care at a health care provider and start of DOTS treatment.
Total delay: time interval between onset of symptoms and start of DOTS treatment.
Statistical analysis
Data from questionnaires were checked before data entry. Data analysis was performed using SPPS, version 11, and Epi-Info 2000. Descriptive statistics were used such as frequency, mean and standard deviation (SD). Comparisons between groups were made using the chi-squared test.
A multivariate logistic regression analysis was performed to analyse the determinants of delay. The median was used as a cut-off to compare patients in 2 groups with low delay (≤ the median) and high delay (> the median value).
Results
Demographic characteristics
More than two-thirds of the 800 newly diagnosed pulmonary TB patients were aged ≤ 35 years old and the mean age was 27.5 years (range 15–95 years). The male to female ratio was 510:290 (1.8). Demographic data showed that 34.6% were illiterate, 44.1% were unemployed or housewives, 67.3% had income that covered their expenses, 34.4% were residing in urban areas, 20.3% in suburbs and 42.9% in rural areas, 75.9% of the patients lived within 5 km of a health facility and 9.3% had to travel more than half an hour to reach a health facility. Half of the patients had a history of current or past smoking, whether cigarettes or nargila (waterpipe) (Table 1).
The majority (80.8%) of patients had “optimal” satisfaction with care and only a few had (19.2%) had “suboptimal” satisfaction. The great majority of patients (94.4%) felt a high degree of stigma attached to TB. Knowledge about TB was poor, with a high proportion of patients (91%) scoring “inadequate” knowledge. Almost all patients (99.6%) had previously heard about TB, mainly from Ministry of Health campaigns (51.7%) or from information from a sick relative or friend (27.0%). There was inadequate knowledge among many patients regarding the presence of a vaccine for the disease (45.3% incorrect), types of drugs (30.2% incorrect) and duration of treatment (14.8% incorrect).
Health-seeking behaviour before diagnosis
When asked about symptoms before diagnosis, cough was reported by almost all the patients (98.3%), followed by fever (76.6%) and weight loss (73.6%). Cough was the main symptom motivating the patients to seek health care (80.4%) (Table 2).
The main health-seeking behaviour after the onset of symptoms was to visit a health care provider (91.9% of patients) (Table 2). The majority of patients sought care at the private sector (79.1%) or public hospitals (17.9%) rather than the NTP centre (1.1%).The health care provider where patients first sought care was most commonly a chest specialist (50.4%), followed by general practitioner (19.3%) and internist (25.8%). The majority of patients visited 1 health care provider before diagnosis, but 41.4% visited more than 1 and up to 5 health care providers (Table 2).
The main reason for delayed care seeking was hoping that symptoms would resolve without treatment (30.3%), while half of patients did not admit to a delay in seeking care (Table 2).
Initial TB diagnosis in the health care system
Despite their initial health-seeking behaviour, almost half the patients were diagnosed with TB in an NTP centre (48.2%). Within the private sector, diagnosis was mainly done by internists (54.7%). The first action after suspecting TB was to request a sputum smear examination and X-ray (95.3%). Diagnosis was rarely based on sputum smear examination only (0.6%) and referral accounted for 2.4% of cases (Table 3).
Reasons for consulting the NTP centre were: free services (30.5%), confidence in getting cured (27.6%), and accessibility (short travel distance) (22.0%). Delay in consulting the NTP centre was attributed to bad experiences (28.4%) and the distance from residence (17.1%).
Determinants of delay
Table 4 summarizes the mean and median delays at all stages between onset of symptoms and start of DOTS treatment.
Patient related diagnostic delay
The mean duration between onset of symptoms and first seeking care was 52.7 days (range 0–426) and the median was 31 days (Table 4).
To establish the determinants of delay due to patient related factors, patients were analysed in 2 groups: delay ≤ the median and > the median. The significant risk factors for delay in seeking care were: inadequate knowledge regarding the disease (1.07-fold increased risk), seeking care at a non-specialized provider (not a health care provider) (5.66-fold increased risk compared with health care provider) and more than 1 health care encounter before diagnosis (1.20-fold increased risk) (Table 5). Diagnostic delay
The mean duration between onset of symptoms and diagnosis was 77.6 days and the median diagnostic delay was 55 days (Table 4). Again, patients were grouped into 2 groups for analysis: ≤ median and > median.
The significant risk factors for diagnostic delay were: older age (1.02-fold increased risk for each year), living far from the health facility (2.20-fold increased risk), high degree of stigma (1.22-fold increased risk), inadequate knowledge regarding the disease (1.07-fold increased risk), seeking care at a non-specialized provider (not a health care provider) (3.22-fold increased risk compared to health care provider); and more than 1 health care encounter before diagnosis (2.14-fold increased risk) (Table 6).
Treatment delay
The mean duration between diagnosis and start of treatment was 2.9 days and the median treatment delay was 1 day from diagnosis (Table 4).
Health care system delay
The mean duration between seeking health care in the health system and start of treatment was 27.6 days and the median health care system delay was 15 days (Table 4).
Total delay
The mean duration between onset of symptoms and start of DOTS treatment was 79.0 days for all patients and the median total delay was 57 days.
Analysing patients above and below the median, the significant risk factors for total delay were living far from the health facility (2.51-fold increased risk), high degree of stigma (1.17-fold increased risk), seeking care at a non-specialized provider (not a health care provider) (3.56-fold increased risk compared to health care provider) and more than 1 health care encounter before diagnosis (2.04-fold increased risk) (Table 7).
Discussion
Delays in diagnosis and treatment of TB can occur at a number of points, from the time a patient develops symptoms until treatment is started on anti-TB drugs. Delays caused by the patient can occur during the process of noticing symptoms, deciding if one is ill, assessing the need for professional care, and overcoming social, personal, and physical barriers to obtaining care from the health care system. Delays in diagnosis can be because the differential diagnosis can broaden or become more focused depending on key pieces of information. For example, a physician who has a high clinical suspicion of TB and a smear-positive sputum result will probably initiate treatment more quickly than one with a low clinical suspicion and a smear-negative result. Furthermore, the clinic may be considering diagnoses other than TB.
The total delay to treatment could be related to the method of estimating the time from onset of symptoms to initiation of treatment, but could also be a true difference in delay to diagnosis. The estimation of the date of onset of symptoms is liable to error, due to recall bias and individual variations in the perception of disease. In addition, what has been defined as the onset of symptoms by the patient could in fact be related to another disease that either coincided with the beginning of TB or had favoured it. Indeed, patients frequently reported symptoms suggesting viral infection or other disease at the onset of disease. The estimate of the delay to treatment therefore lies within a wide range, the limits of which are defined by the occurrence of main events over the year and the individual’s perception of disease.
This study highlights the prolonged delay from the onset of patients’ symptoms until a diagnosis of smear-positive pulmonary TB is made.
In this study the mean patient related diagnostic delay was 52.7 days. The mean patient related diagnostic delay in our study was longer than those reported in similar studies in Egypt, Pakistan and Yemen [8–10], and less than those reported from Somalia [11,12]. Patient related diagnostic delay represented the main part (66%) of the total delay to treatment (79.0 days), whereas the delays in diagnosis and treatment accounted on average for 34% of the total delay. This long patient delay was significantly associated with inadequate knowledge regarding the disease, seeking inital care at non-specialized individuals (not a health care provider) and having more than 1 health care encounter before diagnosis. Residence, sex and age showed no significant association and this may suggest that the most important determining factor for patients taking action about TB is knowledge about the disease; this is similar to what was observed in Yemen [10]. When asked why they had delayed seeking care, 28.4% of patients said that they delayed consulting the national TB control centre due to previous bad experiences and 17.1% because it was far from their residence, 30.3% were hoping that symptoms would resolve without treatment, while half of patients did not admit a delay in seeking care. There is a need to increase awareness of chest symptoms among patients.
The mean diagnostic delay was 77.6 days, which is higher than the mean diagnostic delay in Yemen (57.4 days) and in Egypt (55.9 days) [8,10] and lower than in Pakistan (96.3 days) [9] but similar to Somalia (76.6 days) [11].
A long distance between the patient’s home and the health facility was a significant risk factor for diagnostic delay. Most patients (90.3%) were living within half an hour of the health facility prior to implementation of DOTS, but patients in some rural areas had to travel more than 1 hour to reach a health centre with microscopy facilities. The other significant risk factors for diagnostic delay were: older age, high degree of stigma, inadequate knowledge regarding the disease, seeking care at non-specialized individuals (not a health care provider), and more than 1 health care encounter before diagnosis.
The short treatment delay reflects the standard practice to start treatment as soon as a positive diagnosis is made. The mean treatment delay was 2.9 days: this is longer than in Egypt (1.2 days), Yemen (1.7 days), but less than in Somalia (4.5 days) and Pakistan (4.2 days) [8–11].
The mean health care system delay was 27.6 days. This is longer than in Yemen (20 days), Somalia (19.5 days), but less than in Egypt (33.9 days) and Pakistan (49 days) [8–11].
There are several limitations to this study. First, we were not able to determine the time of onset of symptoms in all patients in the study. Second, the medical records of some patients who sought care from private physicians were not available, which may have resulted in an underestimation of health care system delay in this group. Third, there may have been some recall bias from patients regarding the type, severity and onset of symptoms. Since health workers generally interview patients after they have begun TB treatment, patients may be more likely to report TB rather than non-TB symptoms. Fourth, it can be difficult to differentiate between patient and health care system delays; for example, a patient may have to wait to obtain an appointment with a health care provider. While this may be classified as a patient delay (the patient did not obtain the appointment) it may be more appropriate to classify this as a health care system delay (the patient sought care but it was not immediately available).
This study is valuable for improving the quality of services and strengthening the objectives of disease control; it highlights the importance of improving referral systems and access to diagnostic facilities for TB, at the same time as improving access to treatment if one wishes to reduce transmission of TB in the community. It shows also the importance of increasing awareness of the signs and symptoms of TB in the general population and working closely with health care providers at all levels, including pharmacists, other drug sellers and traditional healers.
The research results were presented at a meeting of the NTP committee and doctors in TB centres, and a series of recommendations based on these results have been made. These recommendations include the provision of on-the-job training to health providers working within and outwith the government health services and the promotion of a concerted effort to increase awareness of signs and symptoms of TB in the general population to encourage self-referral to the health services and thereby increase passive case detection.
Conclusion and recommendations
The long time interval between onset of symptoms and treatment reported in this study was mainly attributed to patient related diagnostic delay rather than delay within the health care system. The main study recommendations are to increase awareness of the community about chest symptoms and the availability of free diagnostic and therapeutic services, educating public and private health care providers about NTP guidelines, and increasing collaboration between both sectors.
Acknowledgements
We would like to express our sincere appreciation to the Syrian Minister of Health for kind approval to carry out the study and for support. Our gratitude is due to the staff of Tropical Diseases Research and Stop TB at the WHO Regional Office (EMRO) who backed this work with interest and support.
This study received technical and financial support from the joint WHO Eastern Mediterranean Region (EMRO), Division of Communicable Diseases (DCD) and the WHO Special Programme for Research and Training in Tropical Diseases (TDR): the EMRO DCD/TDR Small Grants Scheme for Operational Research in Tropical and Other Communicable Diseases.
References
- Stop TB partnership. Annual report 2005. Geneva, World Health Organization, 2006.
- Annual report 2004. Damascus, Ministry of Health, Syrian Arab Republic, 2005.
- Long NH. Longer delays in tuberculosis diagnosis among women. International journal of tuberculosis and lung disease, 1999, 3:388–93.
- Safer MA et al. Determinants of three stages of delay in seeking care at a medical clinic. Medical care, 1979, 17:11–29.
- Acki M, Mori T, Shimao T. Studies on factors influencing patients, doctors and total delay of tuberculosis case-detection in Japan. Bulletin of the International Union against Tuberculosis, 1985, 60:128–30.
- Raj R et al. Factors associated with patient and health system delays in the diagnosis of tuberculosis in south India. International journal of tuberculosis and lung disease, 2002 6:789–95.
- Long NH. Longer delays in tuberculosis diagnosis among women in Vietnam. International journal of tuberculosis and lung disease, 1999, 3:388–93.
- Soliman SS. Case-finding in tuberculosis patients: diagnostic and treatment delays and their determinants. Egypt nationwide. In: Operational research in tropical and other communicable diseases. Final report summaries 2001–2002. Cairo, WHO Regional Office for the Eastern Mediterranean, 2004 (WHO-EM/TDR/007/E).
- Agoatwala M. Case-finding in tuberculosis patients: diagnostic and treatment delays and their determinants. Pakistan, Karachi. In: Operational research in tropical and other communicable diseases. Final report summaries 2001–2002. Cairo, WHO Regional Office for the Eastern Mediterranean, 2004 (WHO-EM/TDR/007/E).
- Al-Absi A. Case-finding in tuberculosis patients: diagnostic and treatment delays and their determinants. Yemen nationwide. In: Operational research in tropical and other communicable diseases. Final report summaries 2001–2002. Cairo, WHO Regional Office for the Eastern Mediterranean, 2004 (WHO-EM/TDR/007/E).
- Abdilahi AI. Case-finding in tuberculosis patients: diagnostic and treatment delays and their determinants. Somalia. In: Operational research in tropical and other communicable diseases. Final report summaries 2001–2002. Cairo, WHO Regional Office for the Eastern Mediterranean, 2004 (WHO-EM/TDR/007/E).
- Wandwalo ER, Morkve O. Delay in tuberculosis case-finding and treatment in Mwanaza, Tanzania. International journal of tuberculosis and lung disease, 2000, 4(2):133–8.
