HIV 感染児における結核治療決定アルゴリズムの検証

Recruitment will be performed at entry in HIV outpatient/inpatient centres of participating hospitals, as well as during the follow up of HIV-infected children on ART. Children will be first pre-screened for TB by attending clinicians and nurses in the context of routine care based on WHO criteria, i.e. with any one of the following symptoms: poor weight gain, fever, current cough and contact with a TB case. If not routinely performed, this pre-screening step will be reinforced and a screening logbook will be set up (if not already in place) to collect aggregated data on the eligibility criteria assessment process. Children pre-screened positive for TB will be further subjected to the study selection criteria by the study nurses.

The PAANTHER algorithm will be used for a TB treatment decision by site clinicians in all children enrolled in the study. The PAANTHER algorithm and prediction score was designed as a guiding tool for empirical TB treatment decision in HIV-infected children with presumptive TB. The proposed score includes:

• history of contact
• suggestive TB symptoms
• tachycardia
• chest radiography
• abdominal ultrasound
• Xpert MTB/RIF Ultra.

A score of >100 is highly predictive of TB. In children with a score >100, anti-TB treatment will be initiated immediately. In children with a score of <100, TB treatment will not be initiated except in those with severe/life-threatening conditions for whom TB treatment decision could be made at the discretion of the site clinician. TB could be definitely ruled out after further assessment and decision made as to what treatment the children should receive in accordance with existing national protocols, along with clinical follow-up.

Proportion of missed TB cases (false negative cases) in children not initiated on treatment as per the PAANTHER TB treatment decision algorithm

The algorithm will be considered completed if a decision to initiate TB treatment has been taken at any step of the algorithm or if TB has been excluded after systematic evaluation, and all steps planned in the algorithm have been implemented.

Proportion of cases considered as unlikely TB by the Expert Committee in those initiated on treatment as per the PAANTHER TB treatment decision algorithm (false positive)

Incidence of morbidity defined as drug-induced toxicity (ART and TB treatment-related), opportunistic infections, and IRIS, with or without TB treatment

Immunologic evolution at 6 months after enrolment defined as CD4 cells gain (absolute count and %)

TB treatment outcome in HIV-infected children initiated on treatment as per the PAANTHER TB treatment decision algorithm defined by weight gain at 6 months (absolute value and percentage of body weight)

Time to Isoniazid Preventive Therapy (IPT) initiation in HIV-infected children not initiated on treatment as per the PAANTHER TB treatment decision algorithm (a)

Proportion of children initiated on IPT in HIV-infected children not initiated on treatment as per the PAANTHER TB treatment decision algorithm (b)

Discrimination (area under the receiver-operating-characteristic curves [AUROC]) and calibration measures of the PAANTHER prediction model including or not MLR and CRP, against the TB composite reference standard as defined by the Expert Committee

Diagnostic performance of Ultra performed on one NPA and one stool sample against mycobacterial culture performed on standard samples in HIV-infected children

Feasibility of NPA and stool samples collection in HIV-infected children defined as the proportion of children with NPA and stool samples collected as per study protocol

Safety of NPA collection defined as proportion of AEs (vomiting, nose bleeding, low oxygen saturation, respiratory distress) occurring during NPA

Tolerability of NPA collection defined as discomfort/pain/distress experienced by the child during NPA, as assessed by the child using the Wong-Baker face scale (in a subset of children).

Scale range: 0 (no hurt) - 5 (hurts worst)

Tolerability of NPA collection defined as discomfort/pain/distress experienced by the child during NPA, as assessed by the parents using the visual analog scale (in a subset of children).

Scale range: 0 (no pain) - 10 (pain as bad as it could possibly be)

Tolerability of NPA collection defined as discomfort/pain/distress experienced by the child during NPA, as assessed by the nurses using the "Face Legs Activity Cry Consolability" behavioral pain scale (in a subset of children).

Total score range: 0 (relaxed and comfortable) - 10 (severe discomfort/pain). Each item of the FLACC scale - Face, Legs, Activity, Cry, Consolability - has 3 possible quotes: 0 or 1 or 2, with a precise description provided to help with the rating. The total score is obtained by adding individual item scores.

false negative rate, i.e. negative predictive value, and corresponding sensitivity of the PAANTHER algorithm in the different subgroups in terms of age, CD4 counts, nutritional statuses, and timings of ART initiation (on/off ART)

false positiverate, i.e. Positive predictive value, and corresponding positivity of the PAANTHER algorithm in the different subgroups in terms of age, CD4 counts, nutritional statuses, and timings of ART initiation (on/off ART)

Incremental cost-effectiveness ratio (ICER)