Background
-
Prospectively--Sustainability refers to the ability, or potential of an entity (local system, health system, organization, etc.) to maintain a function or public good.
-
Retrospectively--Sustainment refers to the verifiable extent to which a public good has measurably evolved over a time.
Method
Identifying a dependent variable
Part 1 - development of the sustainment index
-
b = YT0: baseline value of Y;
-
e = YT1: endline value of Y; and
-
p = YT2: post-project value of Y.
-
p1: Sustainment Index = 2.0. Flawless continuation of progress from T0 to T1 until T2.
-
p2: Sustainment Index = 1.0. Plateau reached after T1.
-
p3: Sustainment Index = 0. Return to baseline conditions, no sustainment.
Natural boundary A
Natural boundary B
-
If C1 = false, the Sustainment Index is not applicable
-
If C2 = true and C3 = true\( \mathrm{Sustainment}\ \mathrm{Index}=1+\left(\frac{p-e}{pmax-e}\right) \)
-
If C2 = false or C3 = false\( \mathrm{Sustainment}\ \mathrm{Index}=1+\left(\frac{p-e/\Delta 2}{e-b/\Delta 1}\right) \)
Part 2 – Testing the sustainment index
-
A face validity validation exercise;
-
Application to a real post-project dataset; and
-
Application to Demographic and Health Surveys (DHS).
Correlation of the sustainment index with an independent practitioners’ panel rating
Application of the sustainment index metric to a published pre- and post-project dataset in Bangladesh
Application of the sustainment index metric to demographic and health survey datasets
Results
Correlation of the sustainment index with an independent practitioners’ panel rating
Application of the sustainment index metric to a published pre- and post-project dataset in Bangladesh
Indicator | 1999 | 2004 | 2009 | Sustainment Index (2009) | |
---|---|---|---|---|---|
Child Health | Complete Immunization | 44% | 91% | 91% | 1.00 |
Vitamin A Supplementation | 37% | 78% | 79% | 1.02 | |
Exclusive breastfeeding | 55% | 72% | 73% | 1.06 | |
Complementary feeding of children 6 to 11 months |
46%
| 64% | 65% | 1.06 | |
Additional feeding and fluids for the sick child | 25% | 44% | 25% | 0.00 | |
Proper child acute respiratory infection (ARI) identification and referral | 24% | 34% | 52% | 2.80 | |
Maternal & Neonatal Health | At least one prenatal consultation during last pregnancy | 59% | 89% | 95% | 1.20 |
At least one tetanus toxoid (TT) dose during last pregnancy | 46% | 89% | 70% | 0.56 | |
Delivery by skilled attendant | 31% | 50% | 59% | 1.47 | |
Delivery in health care facility | 25% | 45% | 57% | 1.60 | |
Immediate breastfeeding | 26% | 57% | 64% | 1.23 |
Application of the sustainment index metric to demographic and health survey datasets
Country | Indicator | T0 (Base %) | T1 (End %) | T2 (Post %) | Sustainment Index |
---|---|---|---|---|---|
Benin | CPR | 7.2 | 6.1 | 7.9 | C1 not met |
Immunization | 59 | 47.1 | 47.6 | C1 not met | |
Burkina Faso | CPR | 4.8 | 8.8 | 15 | 1.89 |
Immunization | 29.3 | 43.9 | 81.3 | 2.46 | |
Cameroon | CPR | 7.1 | 12.5 | 14.4 | 1.30 |
Immunization | 35.8 | 48.2 | 53.2 | 1.35 | |
Cote d’Ivoire | CPR | 4.3 | 7.3 | 12.5 | 1.67 |
Immunization | 37.4 | 50.7 | 50.5 | 0.99 | |
Ethiopia | CPR | 6.3 | 13.9 | 27.3 | 2.47 |
Immunization | 14.3 | 20.4 | 24.3 | 1.53 | |
Ghana | CPR | 13.3 | 18.7 | 16.6 | 0.61 |
Immunization | 62 | 69.4 | 79 | 2.30 | |
Guinea | CPR | 4.2 | 5.7 | 4.6 | 0.37 |
Immunization | 32.2 | 37.2 | 36.5 | 0.88 | |
Kenya | CPR | 31.5 | 31.5 | 39.4 | C1 not met |
Immunization | 59.5 | 51.8 | 68.3 | C1 not met | |
Lesotho | CPR | 35.2 | 45.6 | T2 not available | |
Immunization | 67.8 | 61.7 | C1 not met | ||
Madagascar | CPR | 9.7 | 18.3 | 29.2 | 2.77 |
Immunization | 36.2 | 52.9 | 61.6 | 1.73 | |
Malawi | CPR | 26.1 | 28.1 | 42.2 | 5.70 |
Immunization | 70.1 | 64.4 | 80.9 | C1 not met | |
Mali | CPR | 5.7 | 6.9 | 9.9 | 2.79 |
Immunization | 28.7 | 48.2 | 38.9 | 0.66 | |
Mozambique | CPR | 5.1 | 20.8 | 11.3 | 0.55 |
Immunization | 47.3 | 63.3 | 64.1 | 1.04 | |
Namibia | CPR | 42.6 | 53.4 | 55.3 | 1.21 |
Immunization | 64.8 | 68.7 | 68.4 | 0.91 | |
Niger | CPR | 4.6 | 5 | 12.2 | 25.00 |
Immunization | 18.4 | 29 | 52 | 3.89 | |
Nigeria | CPR | 8.2 | 9.7 | 9.8 | 1.07 |
Immunization | 12.9 | 22.7 | 25.3 | 1.27 | |
Rwanda | CPR | 10.3 | 27.4 | 45.1 | 2.55 |
Immunization | 75.2 | 80.4 | 90.1 | 3.80 | |
Senegal | CPR | 12.1 | 16.1 | 20.3 | 3.10 |
Immunization | 62.8 | 70.2 | 73.7 | 1.95 | |
Tanzania | CPR | 16.9 | 20 | 27.4 | 3.86 |
Immunization | 68.3 | 71.1 | 75.2 | 2.76 | |
Uganda | CPR | 18.2 | 17.9 | 26 | C1 not met |
Immunization | 36.7 | 46.2 | 51.6 | 1.57 | |
Zambia | CPR | 25.3 | 32.7 | 44.8 | 2.17 |
Immunization | 70 | 67.6 | 68.3 | C1 not met | |
Zimbabwe | CPR | 92 | 93 | 89 | −4.60 |
Immunization | 64 | 52.6 | 64.5 | C1 not met |
Discussion
Limitations
-
Initial conditions: as baselines (b) progress toward their optimum value, progress from b to e slows down. This may affect the Sustainment Index in different ways. Caution should be exercised in comparing achieved sustainment under different initial conditions.
-
Natural and secular indicator trend differences: some indicators tend to follow unidirectional changes under strong secular trends, while others are more subject to the vagaries of government commitment, health care financing, development assistance, and innovations, which can affect them upward or downward. The Sustainment Index cannot be interpreted as an absolute value, absent understanding of context, and a theory of change. (We derived no conclusion about causation from Fig. 6, merely observed and described through quantification of shifts in trends.)
Value of the sustainment index for future research
-
Contextual baselines and changes;
-
Major shifts in program investments;
-
Changes in balance of sources of health care financing for different health interventions;
-
Delayed (lagged) effects of capacity strengthening, health system strengthening efforts, policy and political changes on future sustainment indices of relevant indicators;
-
Differences within countries for indicators receiving different types of support;
-
Etc.