SUMMARY
BACKGROUND :
India accounts for 27% of global childhood tuberculosis (TB) burden. Understanding barriers to early diagnosis and treatment in children may improve care and outcomes.
METHODS :
A cross-sectional study was performed among 89 children initiated on anti-TB treatment from a public hospital in Pune during 2016, using a structured questionnaire and hospital records. Health care providers (HCPs) were defined as medical personnel consulted about the child’s TB symptoms. Time-to-treatment initiation (TTI) was defined as the number of days between onset of TB symptoms and anti-TB treatment initiation. Based on Revised National TB Control Programme recommendations, delayed TTI was defined as >28 days.
RESULTS :
Sixty-seven (75%) of 89 enrolled children had significant TTI delays (median 51 days, interquartile range [IQR] 27–86). Sixty-six (74%) children visited 1–8 HCPs in the private sector before approaching the public sector. The median HCP delay was 28 days (IQR 10–75). Bacille Calmette-Gueérin vaccination (aOR10.96, P = 0.04) and loss of appetite (aOR 4.44, P = 0.04) were associated with delayed TTI.
CONCLUSION :
The majority of the children had TTI delays due to delays by HCPs in the private sector. Strengthening HCP competency in TB symptom screening and encouraging early referrals are crucial for rapid scaling up of early treatment initiation in childhood TB.
Keywords: pediatric tuberculosis, TB diagnosis, TB treatment barriers, patient delay, health care provider delay, health care system delay
RÉSUMÉ
CONTEXTE :
L’Inde représente 27% du poids total de la tuberculose (TB) de l’enfant dans le monde. Comprendre les obstacles au diagnostic et au traitement précoces pourrait améliorer la prise en charge et les résultats des enfants atteints de TB.
MÉTHODE :
Une étude transversale a été réalisée parmi 89 enfants mis sous traitement anti-TB dans un hôpital public de Pune en 2016, grâce à un questionnaire structuré et aux registres de l’hôpital. Les prestataires de soins de santé (HCP) ont été définis comme du personnel médical consulté pour des symptômes de TB de l’enfant. Le délai de mise en route du traitement (TTI) a été défini comme le nombre de jours entre l’apparition des symptômes de TB et la mise en route du traitement anti-TB. En se basant sur les recommandations révisées du Programme National de lutte contre la TB, un retard à la mise en route a été défini comme un délai >28 jours.
RÉSULTATS :
Soixante-sept (75%) des 89 enfants enrôlés ont eu un retard significatif de mise en route du traitement (médiane 51 jours ; intervalle interquartile [IQR] 27–86). Soixante-six (74%) enfants ont consulté 1–8 HCP dans le secteur privé avant le système de santé public. Le délai médian des HCP a été de 28 jours (IQR 10–75). La vaccination bacille Calmette-Guérin (OR ajusté [ORa] 10,96 ; P=0,04) et la perte d’appétit (ORa 4,44 ; P = 0,04)ont été associées avec le retard au traitement.
CONCLUSION :
La majorité des enfants a subi un retard au traitement en raison des retards au diagnostic des HCP du secteur privé. Il est crucial de renforcer leurs compètences en matière de dépistage des symptômes de TB et d’encourager une rèfèrence précoce pour parvenir rapidement à une gènèralization de la mise en route précoce du traitement des enfants atteints de TB.
RESUMEN
MARCO DE REFERENCIA:
A la India corresponde el 27% de la carga de morbilidad mundial por tuberculosis (TB) de la niñez. Comprender los obstáculos al diagnóstico temprano y el tratamiento oportuno podría mejorar la atención y los desenlaces clínicos de los niños con TB.
MÉTODOS:
Se llevó a cabo un estudio transversal con 89 niños que iniciaron el tratamiento antituberculoso en un hospital público de Pune durante el 2016, mediante un cuestionario estructurado y el análisis de las historias clínicas. Se definieron los proveedores de atención de salud (HCP) como el personal mèdico consultado por los síntomas indicativos de TB de los niños. Se definió el lapso transcurrido hasta el inicio del tratamiento como el número de días entre el comienzo de los síntomas y el inicio del tratamiento. En conformidad con las recomendaciones del Programa Nacional Revisado de Tuberculosis, se adoptó como límite para un inicio retrasado del tratamiento un lapso superior a 28 días.
RESULADOS:
De los 89 niños inscritos, 67 (75%) presentaron retrasos considerables hasta el inicio del tratamiento (mediana 51 días; amplitud intercuartílica [IQR] 27–86). Sesenta y seis niños (74%) acudieron HCP privado antes de presentarse al sistema público y consultaron de uno a ocho profesionales. La mediana del retraso dependiente de los HCP fue 28 días (IQR 10–75). Se asociaron con un lapso prolongado hasta el inicio del tratamiento la vacunación con el bacille Calmette-Guérin (OR ajustado [ORa] 10,96; P = 0,04) y la pérdida del apetito (ORa 4,44; P = 0,04).
CONCLUSIÓN:
La mayoría de los niños presentó un retraso hasta el inicio del tratamiento antituberculoso, atribuible al HCP privado. Es indispensable fortalecer la competencia de los HCP en la detección de los síntomas de TB y estimular las remisiones tempranas, a fin de ampliar la escala de aplicación del inicio más temprano del tratamiento en los niños con TB.
INDIA ACCOUNTS for 27% of the global childhood tuberculosis (TB) burden.1 The WHO Global TB database (2017) estimated approximately 231 000 Indian children were sick with TB.2 The India TB report observed 1 444 175 TB notifications, of which 6% were pediatric cases.3 Dodd et al. estimated 60 000 childhood TB deaths in India.4 In high TB burden settings, underlying undiagnosed TB may cause 20% of deaths due to pneumonia in children aged under-five.5 Diagnosing pediatric TB is challenging due to non-specific symptoms, paucibaciilary disease and lack of sensitive diagnostic tests. Delays in TB diagnosis and treatment initiation lead to poor outcomes in children.5–8
Data describing delays and barriers to early TB treatment initiation in children are limited.9–13 A study by Beyers et al. reported 4 weeks of TB diagnostic delay.9 The recent Indian study documented 41 days of health system delay.10 Studies related to children with HIV-TB reported longer delays.11,12 A systematic review on barriers to delays in children and youth documented only four pediatric cohort studies.13 The above literature lacked in-depth analysis of pathway of care to treatment initiation in childhood TB.
Better understanding of the reasons for delays in early treatment initiation may inform national efforts to improve care of Indian children with TB. Therefore, we assessed the patient-centered and healthcare system barriers to early treatment initiation among children diagnosed and treated for TB in Pune, an urban setting of high TB incidence in India.
METHODOLOGY
Study setting
A cross-sectional study was conducted at Byramjee Jeejeebhoy Government Medical College (BJGMC) and Sassoon General Hospital (SGH), Pune, India, in 2016, among children initiated on anti-TB treatment from Pediatric and Pulmonary Medicine departments. Estimated TB incidence and prevalence in Pune are 185 and 220 per 100 000 population, respectively.14
India has a mixed healthcare system run by the government and the private sector. Patients can pay for TB diagnosis and care through the private sector. However, the Indian government’s ‘Revised National Tuberculosis Control Programme’ (RNTCP) is implemented through public health facilities, which provide free diagnostic and treatment services. In western India, urban Pune has 11 RNTCP TB units, including SGH. Each TB unit covers a population of 250 000. Patients can visit RNTCP centers directly or can be referred by the private sector. RNTCP guidelines for diagnosis and treatment of pediatric TB15 are followed at SGH, which screens ≥1000 children for TB per year, of which approximately 5–10% are initiated on anti-TB treatment and 7% have microbiologically confirmed disease.
Study population
During January–December 2016, study investigators identified children initiated on anti-TB treatment from the inpatient wards, intensive care units and outpatient departments, by daily review of admission books and case record files. Children aged ≤14 years started on anti-TB treatment were eligible for enrollment as study index cases. Children reported to have anti-TB treatment initiated outside the hospital with no medical records for TB diagnosis such as radiography and tuberculin skin test (TST) were excluded. Written informed consent was provided by the parents/caregivers of eligible cases, for participation in a structured interview. Furthermore, consent included extraction and analysis of data from the child’s hospital case record and review of any health care provider (HCP) prescriptions available from the parents/caregivers.
Parent/caregiver interviews
The parents/caregivers of enrolled pediatric TB patients were interviewed by the principal investigator (CV), to document socio-demographic characteristics and TB symptoms with onset dates using a structured questionnaire adapted from the WHO Global TB report questionnaire.16 If the parent/caregiver did not accurately recall the symptom dates, significant family events were presented to improve the recall. Symptom duration was estimated in days.
Structured parental/caregiver interviews and prescriptions from the HCPs, when available, were used to document the number of HCP visits, contact dates with HCPs, TB contact history, diagnostic tests and treatment for TB by the HCPs. Data on patient delay, health literacy, TB knowledge and stigma, TB contact enquiry in the family and health service factors were also documented.
Hospital case record sheets of the index case
Hospital records were used to document the clinical details, date of the child’s visit to SGH for treatment, diagnostic tests performed, TB diagnosis and anti-TB treatment initiation date.15 Classification of TB as pulmonary TB (PTB)–confirmed, unconfirmed, unlikely, extra-pulmonary TB (EPTB), multidrug resistant TB, presence of co-morbid conditions, HIV-TB co-infection and TB risk factors were also documented.15,17–19
Study outcomes
The primary outcome was time-to-treatment initiation (TTI), defined as number of days between the first onset of any TB symptom and initiation of anti-TB treatment.15 Secondary outcomes included delayed TTI, separate estimates of patient delay, HCP delay and health care system (HCS) delay, and its associated factors (Figure 1).
Figure 1.
Conceptual framework: definitions of delay, showing the steps from patient stage to treatment initiation stage. HCP = health care provider
RNTCP pediatric treatment guidelines
RNTCP pediatric guidelines advise investigation for active-TB among children with fever, cough, weight loss, poor weight gain, and/or a TB contact for .14 days.15 The guidelines recommend an initial 7 day course of antibiotics (without anti-TB activity) before TB investigation, according to the clinical features.15 RNTCP guidelines further recommend that anti-TB treatment be initiated within 7 days of microbiologically confirmed or clinically diagnosed TB.20 Based on these guidelines, we defined TTI to be ≤28 days. HCP delay was defined as duration .>14 days as physicians generally prescribe a 7 day course of antibiotics with a further 7 days required for investigation and treatment initiation (definitions are given in Table 1).15,20
Table 1.
Definitions
| Serial no. | Definitions | |
|---|---|---|
| 1 | Time-to-treatment initiation | The number of days between the onset of any TB symptom and the TB treatment initiation date |
| 2 | Delayed time-to- treatment initiation | Defined as >28 days |
| 3 | Patient delay | The number of days between the first onset of any reported TB symptom (fever, cough, weight loss, loss of appetite, poor weight gain and fatigue) and the first visit of the child to any HCP |
| 4 | HCP | Defined as any medical personnel consulted for the child’s TB symptom(s). These included clinicians (allopathic or indigenous medicine) in the private and public healthcare sectors |
| 5 | HCP delay | The number of days between the first visit to a HCP and the first visit to the SGH RNTCPTB Unit. Defined as >14 days. |
| 6 | Health care system | For this study, this was defined as the SGH RNTCP TB Unit |
| 7 | Health care system delay | Defined as the number of days between the first visit to the health care system and the date of TB diagnosis |
| 8 | TB diagnosis date | The date documented in the medical record that TB was either microbiologically or clinically diagnosed |
| 9 | Diagnostic delay | The number of days between the onset of any TB symptom and the TB diagnosis date |
| 10 | Treatment delay | The number of days between the TB diagnosis date and the TB treatment initiation date |
| 11 | Health service delay | The number of days between the first visit to any HCP and the TB treatment initiation date |
TB = tuberculosis; HCP = health care provider; SGH = Sassoon General Hospital; RNTCP = Revised National Tuberculosis Control Tuberculosis.
Statistical analysis
Statistical analysis were performed using STATA 13.1 for Windows. For descriptive statistics, median and interquartile range (IQR) were calculated. Analysis of categorical variables and treatment initiation delay was performed using Cox regression analysis. When assessing risk factors for delayed TTI, a cut-off point of .>28 days was selected. All statistical tests were two-tailed. Statistical significance was assessed using 95% confidence intervals (CIs) and P < 0.05 was considered significant.
Ethical approval
The study protocol was approved by the Institutional Ethics Committees of BJMC, Pune, India, and Johns Hopkins University, Balitimore, MD, USA to conduct the study at SGH. Written informed consent was provided by the parents/caregivers of eligible children.
RESULTS
Patient’s demographic characteristics
During the study period, 96 children were initiated on anti-TB treatment, of which 89/96 (93%) were enrolled. The parents of three children refused consent and four children initiated on anti-TB treatment outside SGH lacked medical record documentation of TB diagnosis. Interviews were conducted with mothers for 49/89 (55%) children, fathers for 31/89 (35%) and primary caregivers for 9/89 (10%) children. Household demographics and pediatric TB cases clinical characteristics are shown in Tables 2 and 3, respectively. The median age of the children was 7 years (range 0.31–13). TB with malnutrition were found in 41/89 (46%) children. Microbiological confirmed TB were reported in 8/89 (9%) children. Children with PTB and EPTB were 43/89 (48%) and 46/89 (52%), respectively. Of the PTB, 6/43 (14%) children were confirmed and 37/43 (86%) were unconfirmed TB. TB meningitis and HIV-TB co-infection were found in 20/89 (22%) and 16/89 (18%) children, respectively.
Table 2.
Demographic characteristics of pediatric tuberculosis households
| Characteristics | Overall (n = 89) n (%) | Delay TTI (>28 days) (n = 67) n (%) | No delay in TTI (<28 days) (n = 22) n (%) | P value | |
|---|---|---|---|---|---|
| Age of parent/caregiver, years | >0.95 | ||||
| 20–30 | 50 (57) | 38 (58) | 12 (57) | ||
| 31–39 | 28 (32) | 21 (32) | 7 (33) | ||
| ≥40 | 9 (10) | 7 (11) | 2 (10) | ||
| Family type | 0.88 | ||||
| Joint | 28 (66) | 20 (30) | 8 (36) | ||
| Nuclear | 58 (32) | 44 (67) | 14 (64) | ||
| Residence | 0.79 | ||||
| Urban | 66 (74) | 49 (73) | 17 (77) | ||
| Rural | 23 (26) | 18 (27) | 5 (23) | ||
| Slum locality | 0.46 | ||||
| Yes | 34 (38) | 24 (36) | 10 (45) | ||
| No | 55 (62) | 43 (54) | 12 (55) | ||
| Religion | >0.95 | ||||
| Hindu | 71 (80) | 53 (79) | 18 (82) | ||
| Islam | 11 (12) | 8 (12) | 3 (14) | ||
| Christian | 2 (2) | 2 (3) | 0 | ||
| MKS score† | 0.32 | ||||
| I, II, III | 32 (36) | 26 (39) | 6 (27) | ||
| IV, V | 55 (62) | 39 (58) | 16 (73) | ||
| Education of head of household | 0.23 | ||||
| Illiterate | 23 (26) | 16 (24) | 7 (32) | ||
| School education | 51 (58) | 37 (56) | 14 (64) | ||
| College education | 14 (16) | 13 (20) | 1 (5) | ||
| Ventilation in the home | 0.78 | ||||
| Yes | 66 (75) | 50 (76) | 16 (73) | ||
| No | 22 (25) | 16 (24) | 6 (27) | ||
| Smoker in family | >0.95 | ||||
| Yes | 17 (19) | 13 (19) | 4 (18) | ||
| No | 72 (81) | 54 (81) | 18 (82) | ||
| TB in family | 0.81 | ||||
| Yes | 45 (51) | 33 (49) | 12 (55) | ||
| No | 44 (49) | 34 (51) | 10 (45) | ||
| TB treatment status of family member | 0.56 | ||||
| Lost to follow-up | 11 (25) | 7 (21) | 4 (36) | ||
| Completed | 17 (39) | 14 (42) | 3 (27) | ||
| Ongoing | 16 (36) | 12 (36) | 4 (36) | ||
| HIV-infected parent(s) | 0.09 | ||||
| Yes | 14 (22) | 13 (28) | 1 (6) | ||
| No | 50 (78) | 34 (72) | 16 (94) | ||
MKS for socio-economic status in India. (MKS Class I = upper, Class II = upper middle, Class III=middle/lower middle, Class IV = lower/upper lower, Class V = lower).35
TTI = time-to-treatment initiation; MKS = Modified Kuppuswami Score; TB = tuberculosis; HIV = human immunodeficiency virus.
Table 3.
Demographic and clinical characteristics of pediatric tuberculosis patients
| Characteristics | Overall (n = 89) n (%) | Delay TTI (>28 days) (n = 67) n (%) | No delay in TTI (<28 days) (n = 22) n (%) | P value | |
|---|---|---|---|---|---|
| Age, years | 0.91 | ||||
| <1 | 7 (8) | 5 (7) | 2 (9) | ||
| 1–5 | 21 (24) | 15 (22) | 6 (27) | ||
| >5–10 | 44 (49) | 33 (49) | 11 (50) | ||
| >10–14 | 16 (18) | 13 (19) | 3 (14) | ||
| Sex | 0.23 | ||||
| Male | 42 (47) | 29 (43) | 13 (59) | ||
| Female | 47 (53) | 38 (57) | 9 (41) | ||
| Nutritional status* | 0.41 | ||||
| Normal | 3 (4) | 1 (2) | 2 (11) | ||
| MAM | 23 (30) | 19 (33) | 4 (21) | ||
| SAM | 26 (34) | 20 (34) | 6 (32) | ||
| −1 SD to median | 9 (12) | 6 (10) | 3 (16) | ||
| −2 to −1 SD | 16 (21) | 12 (21) | 4 (21) | ||
| BCG-vaccinated | 0.09 | ||||
| Yes | 78 (88) | 61 (91) | 17 (77) | ||
| No | 9 (10) | 4 (6) | 5 (23) | ||
| Previous hospitalization in past 1 year of current TB diagnosis | 0.02† | ||||
| Yes | 31 (35) | 28 (42) | 3 (14) | ||
| No | 57 (65) | 38 (58) | 19 (86) | ||
| Long duration medical condition | 0.1 | ||||
| Yes | 16 (19) | 15 (23) | 1 (5) | ||
| No | 69 (81) | 50 (77) | 19 (95) | ||
| Number of health-seeking encounters | 0.46 | ||||
| ≤2 | 37 (42) | 26 (39) | 11 (50) | ||
| >2 | 52 (58) | 41 (61) | 11 (50) | ||
| Cough >2 weeks | >0.95 | ||||
| Yes | 43 (48) | 32 (48) | 11 (50) | ||
| No | 46 (52) | 35 (52) | 11 (50) | ||
| Fever >2 weeks | 0.28 | ||||
| Yes | 65 (73) | 51 (76) | 14 (64) | ||
| No | 24 (27) | 16 (24) | 8 (36) | ||
| Weight loss | 0.05† | ||||
| Yes | 41 (46) | 35 (52) | 6 (27) | ||
| No | 48 (54) | 32 (48) | 16 (73) | ||
| Loss of appetite | 0.003† | ||||
| Yes | 44 (51) | 39 (61) | 5 (23) | ||
| No | 42 (49) | 25 (39) | 17 (77) | ||
| HIV | 0.06 | ||||
| Positive | 16 (18) | 15 (23) | 1 (5) | ||
| Negative | 72 (82) | 51 (77) | 21 (95) | ||
| Tuberculin skin test | 0.47 | ||||
| Negative | 38 (44) | 30 (46) | 8 (36) | ||
| Positive | 49 (56) | 35 (54) | 14 (64) | ||
| Chest radiograph | 0.04† | ||||
| Normal | 21 (24) | 12 (18) | 9 (41) | ||
| Abnormal‡ | 66 (76) | 53 (82) | 13 (59) | ||
| Ultrasonography abdomen | 0.03† | ||||
| Normal | 42 (64) | 28 (56) | 14 (88) | ||
| Abnormal | 24 (36) | 22 (44) | 2 (12) | ||
| Pulmonary TB | 0.31 | ||||
| Confirmed | 6 (14) | 4 (11) | 2 (25) | ||
| Unconfirmed | 37 (86) | 31 (89) | 6 (75) | ||
| RNTCP treatment category | |||||
| Category I | 84 (95) | 63 (95) | 21 (95) | >0.95 | |
| Category II | 4 (5) | 3 (5) | 1 (5) | ||
WHO classification of nutritional status.
Statistically significant, P < 0.05.
Abnormal chest radiographs consistent with TB—adenopathy, opacity, effusion, cavity, miliary. Read by two treating physicians.
TTI = time-to-treatment initiation; MAM = moderate acute malnutrition; SAM = severe acute malnutrition; SD = standard deviation; BCG = bacille Calmette-Guérin; TB=tuberculosis; HIV = human immunodeficiency virus; RNTCP=Revised National Tuberculosis Control Programme.
Time-to-treatment initiation and delays in the management pathway
The varying time-frames along the patient, HCP, diagnostic and treatment continuum are shown in Figure 2. The median TTI of 51 days (IQR 27–86) and HCP delay of 28 days (IQR 10–75) were prolonged compared to RNTCP guidelines. Sixty-seven of 89 (75%) children had delayed TTI. HCP delay was the major contributor to overall TTI delay. This was due to a median diagnostic delay (DD) of 44 days (IQR 20–78). The median DD when the child directly visited the HCS was 35 days (IQR 12–67) compared to 50 days (IQR 23–87) when the child visited HCPs in the private sector before the HCS. The median delays attributed to the patient (PD, 3 days; IQR 1–14.5), HCS (5 days; IQR 2–10) and treatment (TD, 1 day; IQR 0–2) did not contribute to delays in TTI. The median health service delay (HSD) was 22 days (IQR 10–61.75). The median TTI for PTB and EPTB were 55 days (IQR 34–105) and 46 days (IQR 2–73), respectively. The median TTI for TB meningitis and HIV-TB co-infected children were 35 days (IQR 17–52) and 52 days (IQR 27–79), respectively.
Figure 2.
Box plots of delay categories. PD = number of days between the first onset of any reported TB symptom (fever, cough, weight loss, loss of appetite, poor weight gain, and fatigue) and the first visit of the child to any HCP; HCP delay=number of days between first visit to a HCP and the first visit to the SGH RNTCP TB Unit; HCS delay=the number of days between the first visit to SGH RNTCP TB Unit, i.e., HCS and the TB diagnosis date; DD=the number of days between the onset of any TB symptom and the TB diagnosis date; TD=the number of days between the TB diagnosis date and the TB treatment initiation date; HSD=the number of days between the first visit to any HCP and the TB treatment initiation date; TTI=the number of days between the onset of any TB symptom and the TB treatment initiation date. PD=patient delay; TB=tuberculosis; HCP=health care provider; HCS=health care system; DD=diagnostic delay; TTI time-to-treatment initiation; SGH = Sassoon General Hospital; RNTCP = Revised National Tuberculosis Control Programme; TD = treatment delay; HSD=health service delay.
Factors associated with delayed TTI and HCP delay
The characteristics of household and pediatric TB cases associated with delayed TTI are described in Tables 2 and 3. In univariate analysis, hospitalization in the previous year of current TB diagnosis, bacille Calmette-Guérin (BCG) vaccination, weight loss, fatigue, loss of appetite, abnormal chest radiography and abdominal ultrasonography showed significant association with delayed TTI. In multivariate analysis, BCG vaccination (aOR 10.96, 95%CI 1.36–88.28, P = 0.04) and loss of appetite (aOR 4.44, 95%CI 1.85–23.26, P = 0.04) were associated with delayed TTI (Table 4). BCG vaccination (aOR 28.99, 95%CI 2.47–340.12; P = 0.007) and EPTB (aOR4.95, 95%CI 1.1–22.26; P = 0.04) were independently associated with HCP delay.
Table 4.
Characteristic of pediatric TB cases and HCPs associated with delay in TTI
| Characteristics | n | Delay TTI (>28 days) n (%) | 95% CI | OR (95% CI) | P value | Adjusted OR* (95% CI) | P value |
|---|---|---|---|---|---|---|---|
| Pediatric | |||||||
| Age of child, years, median [IQR] | 7 [3.25–10] | 7.5 [4–10] | — | 1.05 (0.92–1.20) | 0.45 | 1.11 (0.85–1.45) | 0.43 |
| Sex | |||||||
| Male | 42 | 29 (69) | 32–56 | 1 | — | 1 | — |
| Female | 47 | 38 (81) | 44–68 | 1.89 (0.71–5.03) | 0.2 | 2.12 (0.42–10.81) | 0.36 |
| HIV | |||||||
| Negative | 16 | 15 (94) | 14–35 | 1 | — | ||
| Positive | 72 | 51 (71) | 65–86 | 0.16 (0.02–1.30) | 0.09 | ||
| Previous hospitalization in past 1 year of current TB diagnosis | |||||||
| No | 57 | 38 (67) | 45–69 | 1 | — | 1 | — |
| Yes | 31 | 28 (90) | 31–55 | 4.67 (1.26–17.33) | 0.02‡ | 3.09 (0.66–14.42) | 0.06 |
| Long duration medical condition | |||||||
| Yes | 16 | 15 (94) | 14–35 | 1 | — | ||
| No | 69 | 50 (72) | 65–86 | 0.18 (0.02–1.42) | 0.1 | ||
| Number of health-seeking encounters | |||||||
| ≤2 | 37 | 26 (39) | 28–51 | 1 | — | ||
| >2 | 52 | 41 (61) | 49–72 | 1.58 (0.60–4.16) | 0.36 | ||
| TB classification | |||||||
| Pulmonary TB | 43 | 35 (81) | 40–64 | 1 | — | ||
| Extra-pulmonary TB§ | 46 | 32 (70) | 36–60 | 0.5 (0.19–1.41) | 0.2 | ||
| Type of PTB | |||||||
| Confirmed | 6 | 4 (11) | 22–96 | 1 | — | ||
| Unconfirmed | 37 | 31 (89) | 68–94 | 2.58 (0.38–17.43) | 0.33 | ||
| BCG vaccination | |||||||
| No | 9 | 4 (44) | 2–16 | 1 | — | 1 | — |
| Yes | 78 | 61 (78) | 84–98 | 4.49 (1.08–18.56) | 0.04‡ | 10.96 (1.36–88.28) | 0.04‡ |
| Cough >2 weeks | |||||||
| Yes | 43 | 32 (74) | 36–60 | 1 | — | ||
| No | 46 | 35 (76) | 40–64 | 1.09 (0.42–2.87) | 0.86 | ||
| Fever >2 weeks | |||||||
| Yes | 65 | 51 (78) | 64–85 | 1 | — | ||
| No | 24 | 16 (67) | 15–36 | 0.55 (0.19–1.54) | 0.26 | ||
| Weight loss | |||||||
| No | 48 | 32 (67) | 36–60 | 1 | — | 1 | — |
| Yes | 41 | 35 (85) | 40–64 | 2.92 (1.02–8.37) | 0.04‡ | 1.38 (0.34–5.65) | 0.2 |
| Fatigue | |||||||
| No | 58 | 39 (67) | 47–71 | 1 | — | 1 | — |
| Yes | 29 | 26 (90) | 29–50 | 4.22 (1.13–15.72) | 0.03‡ | 1.32 (0.19–9.03) | 0.32 |
| Loss of appetite | |||||||
| No | 42 | 25 (59) | 28–52 | 1 | — | 1 | — |
| Yes | 44 | 39 (89) | 48–72 | 5.30 (1.74–16.20) | 0.003‡ | 4.44 (1.85–23.26) | 0.04‡ |
| Tuberculin skin test | |||||||
| Negative | 38 | 30 (79) | 34–58 | 1 | — | ||
| Positive | 49 | 35 (71) | 41–66 | 0.67 (0.25–1.81) | 0.42 | ||
| Chest radiograph | |||||||
| Normal | 21 | 12 (57) | 11–30 | 1 | — | 1 | — |
| Abnormal¶ | 66 | 53 (80) | 70–89 | 3.06 (1.06–8.79) | 0.04‡ | 0.86 (0.13–5.84) | 0.87 |
| Abdominal ultrasonography | |||||||
| Normal | 42 | 28 (67) | 42–69 | 1 | — | 1 | — |
| Abnormal | 24 | 22 (92) | 31–58 | 5.5 (1.13–26.79) | 0.03‡ | 4.54 (0.56–37.08) | 0.16 |
| Health care providers Education of HCP1 | |||||||
| MD | 15 | 14 (93) | 23–54 | 1 | — | ||
| MBBS | 21 | 16 (76) | 27–59 | 0.23 (0.02–2.2) | 0.2 | ||
| BAMS | 5 | 4 (80) | 4–26 | 0.29 (0.01–5.66) | 0.41 | ||
| Diploma | 5 | 4 (80) | 4–26 | 0.29 (0.01–5.66) | 0.41 | ||
| Education of HCP2 | |||||||
| MD | 22 | 15 (68) | 30–65 | 1 | — | ||
| MBBS | 12 | 9 (75) | 15–47 | 1.4 (0.29–6.83) | 0.68 | ||
| BAMS | 4 | 4 (100) | 5–30 | — | — | ||
| Diploma | 4 | 4 (100) | 5–30 | — | — | ||
| Inquiry about TB by HCP1 | |||||||
| Yes | 7 | 6 (86) | 5–24 | 1 | — | ||
| No | 58 | 45 (78) | 76–95 | 0.58 (0.06–5.23) | 0.62 | ||
| Visit to health system | |||||||
| Private | 63 | 47 (75) | 58–80 | 1 | — | ||
| Public | 26 | 20 (77) | 20–42 | 1.13 (0.39–3.32) | 0.82 | ||
Adjusted for significant variables of univariate analysis.
Statistically significant, P < 0.05.
Tuberculous meningitis, neurotuberculoma, abdominal tuberculosis, TB lymphadenitis, pleural effusion, pericardial effusion, skin TB.
Abnormal chest radiographs consistent with tuberculosis: adenopathy, opacity, effusion, cavity, miliary. Read by two treating physicians.
TB=tuberculosis; HCP=health care provider; TTI=time-to-treatment initiation; OR=odds ratio; CI=confidence intervals; IQR=interquartile range; HIV=human immunodeficiency virus; PTB=pulmonary TB; BCG=bacille Calmette-Guérin; MD=Doctor in Medicine; MBBS=Bachelor of Medicine Bachelor of Surgery; BAMS=Bachelor of Ayurveda Medicine and Surgery.
Perceived barriers by parents/caregivers to early TTI
The private sector as first point of healthcare was reported by 63/89 (71%) patients. Parents/caregivers reported consulting up to eight HCPs in the private sector, before seeking care at HCS. Forty-four of 89 parents (49%) trusted private sector HCPs. More than three HCPs were visited by 41/89 (46%) patients. Of the 67/89 (75%) children with documented delay, 46/67 (90%) parents/caregivers felt there was delay in TB diagnosis for their child (P < 0.001) (Table 5). Perceived barriers by parents/caregivers to early TTI are described in Table 5. Knowledge on TB as a contagious disease was lacking in 44/89 (49%) and the fact that it was curable was lacking in 40/89 (45%) parents/caregivers.
Table 5.
Perceived barriers by parent/caregiver for delay in early TTI
| Barriers to early treatment initiation | Total (n = 89) n/N (%) | Delay TTI (>28 days) (n = 67) n/N (%) | No delay in TTI (<28 days) (n = 22) n/N (%) | P value | |
|---|---|---|---|---|---|
| Delays as perceived by the parent/caregiver (n = 89) | <0.001* | ||||
| Yes | 51/89 (57) | 46/51 (90) | 5/51 (10) | ||
| No | 34/89 (38) | 17/34 (50) | 17/34 (50) | ||
| Unsure | 4/89 (4) | 4/4 (100) | 0 | ||
| Perceived barriers by parents/caregivers† | (n = 51) | (n = 46) | (n = 5) | ||
| Patient-related delays | |||||
| Unaware symptoms may be related to TB | 32/51 (62) | 30/32 (94) | 2/32 (6) | 0.06 | |
| No TB in family | 9/51 (17) | 8/9 (89) | 1/9 (11) | >0.95 | |
| Economic constraints | 3/51 (6) | 1/3 (33) | 2/3 (67) | 0.05 | |
| Fear of what would be found on diagnosis | 4/51 (8) | 3/4 (75) | 1/4 (25) | 0.48 | |
| Hoped their symptoms would go away on their own (denial and concealment) | 7/51 (14) | 7/7 (100) | 0 | 0.58 | |
| Fear of social isolation | 2/51 (4) | 2/2 (100) | 0 | >0.95 | |
| Health care provider-related delays: private sector (as perceived by parents/caregivers) | |||||
| Poor quality of health services: lack of inquiry of TB symptom screen, investigations and TB contacts | 21/51 (41) | 20/21 (95) | 1/21 (5) | 0.14 | |
| Inadequate staff attitude | 6/51(12) | 6/6 (100) | 0 | 0.58 | |
Statistically significant, P < 0.05.
Reasons are listed based on the frequency and the multiple answers reported by the parents TTI = time-to-treatment initiation; TB = tuberculosis.
DISCUSSION
This study observed significant delays in TB treatment initiation in Indian children. We assessed TTI as a marker of early detection of pediatric TB, as it assists in preventing poor outcomes21 and also plays a significant role in global programs for TB elimination. Few previous studies have focused on TTI in pediatric cohorts.9,10,13,22 We observed a median TTI of 51 days, which is similar to another study from India (median 52 days), where delay was evaluated as quartile groups.10 Other studies conducted in South Africa9 and the United Kingdom22 found shorter median diagnostic delay (range 36–37 days) compared to our study. These differences could be explained by the strategy of including private sector health visits as part of the TTI delay in our study. Of note, we found the TTI delay was mainly due to a median of 28 days between the initial HCP contact and visit to the HCS.
Our study observed that when Indian children have symptoms, their parents/caregivers quickly ensure they are seen by a clinician (PD median 3 days). When the children access the public HCS, the time to TB diagnosis (median 5 days) and subsequent TB treatment initiation (median 1 day) is relatively short. The HCP delay may be due to multiple HCP visits by the patient in the private sector and late referrals to the public HCS. Our data suggest that TTI delay occurs mainly due to private sector HCP delay. This could be improved by strengthening the private sector HCP’s knowledge to refer children with suspected TB symptoms promptly to the HCS for TB diagnosis and treatment initiation. Standardization of definitions for HCP delay in the private sector may help to bridge the gap as numerous studies applied different definitions.23–25
Mistry et al. observed a 65 day delay in TTI among adults and children with TB.26 The study found major contributor to delays were the diagnostic process (42 days), similar to our study. The low proportion of children aged under-five in our study was consistent with other studies.10,27,28 This could be due to missed cases in the community due to diagnostic difficulty.5,6 Half of the enrolled children were malnourished, where TB diagnosis is a challenge due to false-negative TSTand atypical presentation.29 In addition, TB symptoms may resemble diseases that are common in this setting, such as malaria, asthma and pneumonia. The symptomatology of our cohort was similar to studies from Vietnam and Nigeria.27,28 Our finding of loss of appetite associated with delays, emphasizes the need for reassessing validation of symptom-based approaches for diagnosing childhood TB in high-burden settings.30 Pediatric TB guidelines use TB symptoms, diagnostic tests and TB contact as clues to diagnosis,15 due to poor bacteriological confirmation.5 Compared to published literature, our study found low bacteriologically confirmed TB cases.7,27,31 This may be attributed to high rate of EPTB in our study. HCP delays were associated with EPTB which reiterates the diagnostic challenges faced in these children. Our study found that delay in TTI was strongly associated with receipt of BCG vaccine. This may be due to clinicians assuming that BCG is fully protective against TB in high-burden settings.32 This possibility warrants further study. In spite of timely healthcare consultation, our study found parents/caregivers felt there was delay in TB diagnosis similar to the Peru study.12 Compared to our study, previous pediatric studies lacked analysis of household, pediatric and HCP’s characteristics and TTI.9–13,22,26
A systematic review of children and youth in the sub-Saharan African region highlighted cost, health system infrastructure and health-seeking behaviors as barriers to TB treatment initiation.13 Our study found parental/primary caregivers preference for the private sector, similar to previous studies.25,33 TB notification by the HCPs in the private sector is mandated by the Indian government,34 but recognition of TB symptoms in children by HCPs in the private sector, is lacking. Capacity building of HCPs in the private sector for recognition of TB symptoms in children would be improved by frequent mandatory training programs. Also, the development, monitoring and surveillance of an HCP training database by the RNTCP would help the program to achieve its goal. RNTCP should identify HCPs around a public health center for linkage and referral networks. Building strategic public private partnerships for early pediatric TB diagnosis would further help to bridge this gap.
Our study contributes to a better understanding of gaps in childhood TB treatment initiation in an urban Indian setting. A strength of our study lay in enrolling newly diagnosed children, thereby minimizing recall bias. Furthermore, assessing the data from the medical records and parental/caregiver interviews improved the validity.
However, the use of a small cohort from a single RNTCP TB unit of a public sector provider with analysis of only 67 children with delayed TTI limits the conclusions we can draw. The lack of a complete understanding of the private HCP’s prescriptions also adds to the limitations of our study. The small number of confirmed TB cases in our study may result in an overestimation of the TTI.
Barriers to early TB diagnosis in the private sector, delayed HCP’s referrals from private to public sector and late access of public sector by the parents need to be addressed. Future studies are required to assess the impact of delayed TTI on TB outcomes, schooling and quality of life.
CONCLUSION AND RECOMMENDATIONS
We found delays in TTI for pediatric TB cases among 75% of the study cohort. Strengthening HCP awareness regarding the importance of the TB symptom screen and encouraging early referrals are crucial for rapid scaling up of early treatment initiation in childhood TB. To be successful in its campaign to eliminate TB in India, the RNTCP needs to work closely with the private sector to reduce the number of HCPs consulted before a diagnosis can be reached – with regular training and resources for a functioning pediatric TB referral system.
Acknowledgements
The authors thank the patients and staff of Sassoon General Hospital, Pune, India, for their co-operation and valuable inputs. This study is supported by the Byramjee Jeejeebhoy Government Medical College (BJGMC) Johns Hopkins University (JHU) Fogarty HIV-TB Training Program funded by the Fogarty International Center, National Institutes of Health (NIH) (grant # D43TW009574). This analysis and manuscript preparation has also been funded in part by support of cTRIUMPh: Cohort for TB Research by the Indo-US Medical Partnership with Federal funds from the Government of India’s (GOI) Department of Biotechnology (DBT), the Indian Council of Medical Research (ICMR), the US NIH, National Institute of Allergy and Infectious Diseases (NIAID), Office of AIDS Research (OAR), and distributed in part by CRDF Global. The authors also thank S Gaikwad and V Belgaumkar, faculties of BJGMC and SGH, Pune, India, for their valuable inputs.
Footnotes
Disclaimer: The contents of this publication are solely the responsibility of the authors and do not represent the official views of the NIH, BJGMC, JHU, DBT, the ICMR, the NIH, or CRDF Global. Any mention of trade names, commercial projects, or organizations does not imply endorsement by any of the sponsoring organizations.
Conflicts of interest: none declared.
References
- 1.Dodd PJ, Gardiner E, Coghlan R, Seddon JA. Burden of children tuberculosis in 22 high-burden countries: a mathematical modelling study. Lancet Glob Health 2014; 2: 453–459. [DOI] [PubMed] [Google Scholar]
- 2.WHO’s Global TB database. http://www.who.int/tb/data/en/Accessed on January 2019.
- 3.Ministry of Health and Family Welfare, Government of India India TB Report 2018. New Delhi, India: MoHFW, 2018. https://tbcindia.gov.in/showfile.php?lid=3314. Accessed August 2019. [Google Scholar]
- 4.Dodd PJ, Yuen CM, Sismanidis C, Seddon JA, Jenkins HE. The global burden of tuberculosis mortality in children: a mathematical modelling study. Lancet Glob Health 2017; 5: e898–e906. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Oliwa JN, Karumbi JM, Marais BJ, Madhi SA, Graham SM. Tuberculosis as a cause or comorbidity of childhood pneumonia in tuberculosis-endemic areas: a systematic review. Lancet Respir Med 2015; 3: 235–243. [DOI] [PubMed] [Google Scholar]
- 6.United Nations Childrens’ Fund Change the game: an agenda for action on childhood tuberculosis. New York, NY, USA: UNICEF, 2018. https://data.unicef.org/wp-content/uploads/2018/03/TB-Advocacy-Brochure-Final-3_21-high-Res.pdf Accessed August 2019. [Google Scholar]
- 7.Lolekha R, Anuwatnonthakate A, Nateniyom S, et al. Childhood TB epidemiology and treatment outcomes in Thailand: a TB active surveillance network, 2004 to 2006. BMC Infect Dis 2008; 8: 94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Mtabho CM, Irongo CF, Boeree MJ, Aarnoutse RE, Kibiki GS. Childhood tuberculosis in the Kilimanjaro region: lessons from and for the TB programme. Trop Med Int Health 2010; 15(5): 496–501. [DOI] [PubMed] [Google Scholar]
- 9.Beyers N, Gie RP, Schaaf HS, et al. Delay in the diagnosis, notification and initiation of treatment and compliance in children with tuberculosis. Tubercle Lung Dis 1994; 75: 260–265. [DOI] [PubMed] [Google Scholar]
- 10.Kalra A Care seeking and treatment related delay among childhood tuberculosis patients in Delhi, India. Int J Tuberc Lung Dis 2017; 21: 645–650. [DOI] [PubMed] [Google Scholar]
- 11.Felicitee N, Ntoto NKM, Roger D, et al. Pulmonary tuberculosis in children with HIV infection: from symptoms to diagnosis and treatment in a pediatric center in Yaoundé-Cameroon. J HIV Retrovirus 2016, 2: 1. [Google Scholar]
- 12.Paz-Soldan VA, Alban RE, Dimos Jones C, Powell AR, Oberhelman RA. Patient reported delays in seeking treatment for tuberculosis among adult and pediatric TB patients and TB patients co-infected with HIV in Lima, Peru: a qualitative study. Front Public Health 2014; 2: 281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Sullivan BJ, Esmaili BE, Cunningham CK. Barriers to initiating tuberculosis treatment in sub-Saharan Africa: a systematic review focused on children and youth. Glob Health Action 2017; 10(1): 1290317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Central TB Division, Directorate General of Health Services, Ministry of Health and Family Welfare, Government of India TB India 2014 Revised National TB Control Programme Annual Status Report. New Delhi, India: GOI, 2014. https://tbcindia.gov.in/showfile.php?lid=3142 Accessed August 2019. [Google Scholar]
- 15.Revised National Tuberculosis Control Programme RNTCP guidelines: National guidelines on diagnosis and treatment of pediatric tuberculosis, 2012. New Delhi, India: RNTCP, 2012. https://tbcindia.gov.in/showfile.php?lid=2904. Accessed August 2019. [Google Scholar]
- 16.World Health Organization. Diagnostic and treatment delay in tuberculosis: an in-depth analysis of the health-seeking behaviour of patients and health system response in seven countries of the Eastern Mediterranean Region WHO-EM/TDR/009/E Geneva, Switzerland: WHO, 2006. http://applications.emro.who.int/dsaf/dsa710.pdf Accessed August 2019. [Google Scholar]
- 17.Graham SM, Cuevas LE, Jean-Philippe P, et al. Clinical case definitions for classification of intrathoracic tuberculosis in children: an update. Clin Infect Dis 2015; 61(Suppl 3): S179–S187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Marais S, Thwaites GE, Schoeman JF, et al. Tuberculous meningitis: a uniform case definition for use in clinical research. Lancet Infect Dis 2010; 10: 803–812. [DOI] [PubMed] [Google Scholar]
- 19.Revised National Tuberculosis Control Programme Guidelines on programmatic management of drug resistant tuberculosis (PMDT) in India. New Delhi, India: RNTCP, 2012. https://tbcindia.gov.in/showfile.php?lid=3155 Accessed August 2019. [Google Scholar]
- 20.Central TB Division (CTD), Directorate General of Health Services, Ministry of Health and Family Welfare, Government of India Strategy document for the supervision and monitoring of the Revised National Tuberculosis Control Programme. Delhi, India: GOI, 2005: p 4 http://health.bih.nic.in/Docs/Guidelines/Guidelines-TB-Control.pdf Accessed August 2019. [Google Scholar]
- 21.World Health Organization. Global tuberculosis report, 2016 WHO/HTM/TB/2016.13 Geneva, Switzerland: WHO, 2016. [Google Scholar]
- 22.Abubakar I, Laundy MT, French CE, Shingadia D. Epidemiology and treatment outcome of childhood tuberculosis in England and Wales: 1999–2006. Arch Dis Child 2008; 93: 1017–1021. [DOI] [PubMed] [Google Scholar]
- 23.Zerbini E, Chirico MC, Salvadores B, Amigot B, Estrada S, Algorry G. Delay in tuberculosis diagnosis and treatment in four provinces of Argentina. Int J Tuberc Lung Dis 2008, 12: 63–68. [PubMed] [Google Scholar]
- 24.Storla DG, Yimer S, Bjune GA. A systematic review of delay in the diagnosis and treatment of tuberculosis. BMC Public Health 2008; 8: 15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sreeramareddy CT, Qin ZZ, Satyanarayana S, Subbaraman R, Pai M. Delays in diagnosis and treatment of pulmonary tuberculosis in India: a systematic review. Int J Tuberc Lung Dis 2014; 18(3): 255–266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mistry N, Rangan S, Dholakia Y, Lobo E, Shah S, Patil A. Durations and delays in care seeking, diagnosis and treatment initiation in uncomplicated pulmonary tuberculosis patients in Mumbai, India. PLoS One 2016; 11: e0152287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Blount RJ, Tran B, Jarlsberg LG, et al. Childhood tuberculosis in Northern Viet Nam: a review of 103 cases. PLoS One 2014; 9(5): e97267. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ogbudebe CL, Adepoju V, Ekerete-Udofia C, et al. Childhood tuberculosis in Nigeria: disease presentation and treatment outcomes. Health Serv Insights 2018; 11: 1178632918757490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Chisti MJ, Ahmed T, Pietroni MA, et al. Pulmonary tuberculosis in severely-malnourished or HIV-infected children with pneumonia: a review. J Health Popul Nutr 2013; 31: 308–313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Marais BJ, Gie RP, Obihara CC, Hesseling AC, Schaaf HS, Beyers N. Well defined symptoms are of value in the diagnosis of childhood pulmonary tuberculosis. Arch Dis Child 2005; 90(11): 1162–1165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Turul O, Kazanci S, Gonen I, Aydogmus C, Karaoglan E, Siraneci R. Paediatric tuberculosis at a referral hospital in Istanbul: analysis of 250 cases. BioMed Res Int 2016; 6896279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Roy A, Eisenhut M, Harris RJ, et al. Effect of BCG vaccination against Mycobacterium tuberculosis infection in children: systematic review and meta-analysis. BMJ 2014; 349: g4643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Atre S, Murray M. Management and control of multidrug-resistant tuberculosis (MDR-TB): addressing policy needs for India. J Public Health Pol 2016; 37: 277–299. [DOI] [PubMed] [Google Scholar]
- 34.Central TB Division (CTD), Directorate General of Health Services, Ministry of Health and Family Welfare, Government of India. Gazette on mandatory TB notification New Delhi, India: GOI, 2018. https://tbcindia.gov.in/showfile.php?lid=3322 Accessed August 2019. [Google Scholar]
- 35.Guru Raj MS, Shilpa S, Maheshwaran R. Revised socioeconomic status scale for urban and rural India–revision for 2015. Socioeconomica 2015; 4(7): 167–174. [Google Scholar]