OIG Investigation of key HIV behavioural survey in Guinea finds falsified data and costs
In an investigation of a Global Fund HIV grant to Guinea, the Office of the Inspector General (OIG) has found that in 2015, a local NGO called SIDALERTE falsified survey data and survey costs, which misrepresented the programme’s progress.
The OIG’s investigation report was made public on 30 October 2018.
The total value of SIDALERTE’s contract, $114,366, is the amount the OIG proposes the Global Fund recover in non-compliant expenditures – but the implications of the data falsification reach beyond the financial costs.
SIDALERTE was contracted in 2015 by Guinea’s National AIDS Council (Comité National de Lutte contre le SIDA; CNLS), the Principal Recipient of an HIV grant that ended in June 2018, to conduct an Integrated Biological and Behavioral Surveillance (IBBS) survey. The survey was intended to update core HIV behaviour and prevalence indicators among at-risk population groups, and to provide an information base for the mid-term evaluation of Guinea’s 2013-2017 national AIDS strategic framework.
The investigation found that about one quarter (2,306 of the survey’s 9,740 participants, or 23.7%) were fictitious, and had been created to cover region-specific gaps in numbers of survey participants and HIV tests that fell short of targets. (The technical oversight team cited the then-ongoing Ebola crisis as having hindered survey participation rates.) The 2015 IBBS surveyed individuals from eight at-risk groups: youth, miners, men in uniform, female sex workers, prisoners, fishermen, truck drivers, and TB patients. The replications were highest in the miners and youth groups – 48% and 38% respectively.
The falsified data were created by replicating the demographic characteristics and survey responses of 1,176 other (real) participants, and the numbers of blood tests inflated by more than 50%.
Non-compliant expenditures totalled $114,366, which represents the full value of the SIDALERTE contract. The OIG proposes that the Secretariat seek recovery of the entire amount.
An important, less tangible finding of the OIG’s that is understated in the report’s top-line summary is that the investigation “also found substantial issues with data quality”. For 83% of the survey questionnaires analyzed, the OIG could not match response data to participant responses in the survey’s database. Reported HIV prevalence rates did not reflect (nor could be linked to) the prevalence results recorded by the laboratory technicians who had conducted the HIV tests. In addition, the accuracy of the initial HIV rapid test results had not been independently validated, though this had been planned and budgeted for.
Origins of the OIG investigation
The wrongdoing began in May 2015, but the OIG was only alerted to it in February 2017, when the Secretariat signaled to the OIG the possibility of fraud by SIDALERTE. The NGO filed a claim for reimbursement of some line items from the budget for the 2015 IBBS. During 2017, the OIG conducted two missions to Guinea, obtaining copies of the IBBS database of survey participants and responses, as well as copies of “a large sample” of the hard-copy questionnaires from the field, lab technicians’ HIV test registers, and financial and other records.
The OIG identified five main findings in this investigation – in brief:
- A quarter of the survey’s recorded participants were fictitious, and were exact replications of other survey participants’ characteristics and responses. The total of 9,740 individuals.
- Survey participants reported demographics and survey responses could not be matched to the underlying questionnaires.
- SIDALERTE inflated the reported number of blood samples collected and tested for HIV by over 50% overall, and falsified the reported HIV prevalence rates (the reported rates were not supported by the underlying HIV test evidence at the laboratory level); the actual prevalence rates recorded in the laboratory test registers were higher than reported for seven of the eight target groups in the survey, and more than twice as high for four of the groups. In addition, independent quality assurance blood tests that had been budgeted for had not been performed.
- SIDALERTE submitted falsified documents with inflated amounts in support of survey expenses paid (AMA 1 and 2 relate to this finding).
- The Principal Recipient did not provide adequate oversight activities (AMAs 3, 4 and 5 relate to this finding).
Types of wrongdoing, and agreed management actions
The report identifies “fraud” as the single type of wrongdoing found in this investigation.
Based on the report findings, the Secretariat will ensure the carrying out of five agreed management actions, summarized below. (For a full table of AMAs, please see page 17 of the OIG report.)
|Agreed Management Action||Target date|
|1. The full amount of the SIDALERTE contract, $114,366, be recovered (the entire contract amount is considered non-compliant).||30 September 2019|
|2. SIDALERTE is considered for sanctions; it and its principals are prevented from future participation as a supplier in Global Fund-financed programs.||31 March 2019|
|3. Guinea’s Country Coordinating Mechanism and in-country partners are informed of the investigations’ findings and nullify the 2015 IBBS results.||30 November 2018|
|4. Secretariat will retain an appropriate service provider to conduct a review of the Guinea 2017 IBBS recently conducted by an international NGO, to provide assurance of the survey’s integrity and accuracy.||31 March 2019|
|5. Secretariat will develop operational guidance describing policy and guidance for IBBS and similar surveys, to ensure accuracy and address potential risks.||
31 March 2019
(See the full table of AMAs on page 17 of the OIG report.)
The Global Fund has signed more than $330 million in grants to Guinea, and has disbursed $214 million, since 2003. At the moment there are three active Global fund grants in the country (see Table below) with a total value of $199.4 million. Two of these grants, an HIV and a TB-HIV grant, were signed in 2018.
The 2015 IBBS survey which was the focus of this investigation was financed from a grant that was active from 2012 and closed in June 2018 (GIN-H-CNLS).
Guinea is a ‘core’ Global Fund country, and is also designated a Challenging Operating Environment, and has an Additional Safeguard Policy. With a population of 12.6 million people, 55% of whom live below the poverty line, Guinea is ranked 183 out of 188 countries in the UNDP human development index report.
|Principal Recipient||Grant component||Grant||Signed amount (US$)|
|Ministry of Public Health||HIV||GIN-H-MOH||33,573,749|
|Catholic Relief Services||Malaria||GIN-M-CRS||151,257,623|
Previously identified issues
The OIG has previously published one other investigation report concerning Guinea, in March 2015. That investigation had uncovered “serious misappropriation and fraud” concerning 22 of 26 sub-sub-recipients that had been working on Global Fund HIV programs in the country between 2008 and 2010. The investigation uncovered a total of $416,183 in misappropriated or unsupported non-compliant expenditures.
In 2010, the Secretariat stopped using the sub-sub-recipients who had been implicated. In 2012, the Secretariat invoked the Additional Safeguard Policy, and in 2014, appointed the National AIDS Council (Conseil National de Lutte contre le SIDA) as PR, to replace the Ministry of Health.
Global Fund Response
Asked for comment on this OIG investigation report into the HIV data falsification issue, Global Fund Director of Communications Seth Faison told Aidspan, via email: “We take this issue extremely seriously. The Global Fund is increasingly attuned to the importance of data integrity and data quality, essential building blocks of working effectively in global health. We are actively strengthening our work on data, and we will continue to be on the lookout for any obstacles.”
The full report of the OIG investigation into data falsification within Guinea’s Global Fund-supported HIV grant is accessible on the Global Fund website, as well as the March 2015 OIG Investigation report, and the 2017 OIG audit report.