Table 1.

Summary of findings from evaluation of DETECT Child TB Project [14,15].

Project Outcome	Data Analysis #	External Evaluation Findings
Decentralisation and Strengthening of Child TB Services
Proportion of all diagnosed child TB cases by health facility level *	Baseline: 96% at level V, 3% at level IV and 1% at level III By end implementation (Q4 2016): 50% at level V, 21% at level IV and 29% at level III	The DETECT model demonstrated that child TB services can be successfully decentralised with the greatest increase in detection occurring at the level III facility in both districts.
Health worker knowledge, Wakiso District	Average (range) test score Baseline: 40 (28–70)% Implementation: 75 (64–89)%	The ability and confidence of health workers in peripheral health facilities to diagnose TB in children was improved and the numbers of of unnecessary referrals were reduced.
Health worker knowledge, Kabarole District	Average (range) test score Baseline: 45 (32–64)% Implementation: 78 (68–96)%
Number of functional TB basic management units for diagnosis and treatment by district	Baseline: 24 in Kabarole and 41 in Wakiso District By end implementation: 30 in Kabarole and 46 in Wakiso District	Repair of non-functional microscopes and re-training of laboratory personnel benefit TB services and detection for all ages. Improved laboratory and drug supplies with reduced stock-outs.
Changes to TB Case Detection and Treatment Outcomes
Caseload of child TB and as a proportion of total TB notifications	139% increase in 0–14 years Baseline: 271, 8.8% Implementation: 647, 14.9%	The majority of respondents from the focus group discussions and in-depth interviews felt that the project had improved delivery of TB services, as it had accomplished the following: built capacity of health workers; increased detection of child TB cases; increased confidence in child TB management; strengthened facility and community household contact-tracing activities; built and increased community trust in healthcare workers through integration; contributed to a reduction in deaths in children. No respondents thought that the project was a burden to the community.
TB cases in young children, proportion of all child TB cases	Increase in <5 years ageBaseline: 99, 36.5% Implementation: 324, 50.1%
Bacteriologically confirmed (BC) TB cases in children	61% increase in BC cases detected, but proportion with BC remained low in children. Baseline: 44 (9 young) children 0–14 years: 44/171 = 16% BC <5 years: 9/99 = 9% BC Implementation: 71 (16 young) children 0–14 years: 71/647 = 11% BC <5 years: 16/324 = 5% BC
Cases of TB in older adolescents and adults	32% increase Baseline: 2805 Implementation: 3693
Treatment success, cure or treatment complete	Significant improvement Baseline: 65% Implementation: 81%
Died or treatment failure	Reduction in poor outcomes: Baseline: 15% Implementation: 4%
Household Child Contact Screening and Management ^
Training and support of community health workers to implement	178 (target was 168) community health workers received training	At least two per facility were trained over 2 days, provided with job aides and recording tools, linked with the facility-based TB focal person for ongoing mentorship and supervision.
Households screened	1617 households with 2270 child contacts	The project increased screening of households of smear-positive TB patients by 142%.
Child contacts with positive symptom screen	602 (27%) of 2270 child contacts were symptomatic	Screening symptoms used were cough, weight loss or poor weight gain, fever or lethargy/reduced playfulness.
Child contacts evaluated for TB disease	486 (81%) of 602 symptomatic child contacts	19% of symptomatic child contacts did not present to the health facility for further evaluation.
Child contacts diagnosed with TB	55 child TB cases detected	Lower numbers than expected for overall case detection—2.4% of all child contacts. Of symptomatic child contacts who presented to the facility for evaluation, 11% diagnosed with TB.
Child contacts eligible for TPT	910 young child contacts without active TB identified	Challenges with availability of isoniazid-alone preparation for TPT at beginning of project.
Eligible child contacts who initiated IPT	670 or 77% of 910 eligible	Remarkable improvements in IPT uptake noted in both districts over time but lower than the 90% target.
Child contacts who completed IPT	569 or 85% of 670 who commenced IPT	Although short of the 90% target, a high rate of completion.

TB: tuberculosis; TPT: tuberculosis preventive treatment; IPT: isoniazid preventive treatment. # Comparison was between two separate 18-month periods: baseline or pre-implementation was January 2014–June 2015; and implementation was July 2015–December 2016. * Level V: regional referral hospital for the district; level IV: secondary level health facility with laboratory and inpatient care; level III: primary level health facility with sputum smear microscopy available. ^ Routine household screening by community health workers was introduced by the project, and so there were no baseline programmatic data.