Skip to end of metadata
Go to start of metadata


There are several program recommendations which have emanated from the Lost-To-Follow-UP (LTFU) project. These recommendations have been put into three broad categories: Clinical management, Database management and Records management.

Please note that these are just the preliminary recommendations as the full results of the evaluation are yet to be published. For more information, please contact Daniel Ochieng from Outreach.

Clinical Management Recommendations:

  • A new system or strengthening the current systems (and SOP) needs to be put in place to prevent patients from bypassing routine AMPATH processes including:
    • Use the correct version of forms/ mark the forms correctly (e.g. antenatal care and post natal care).
    • Need to follow protocols/SOPs which have been delivered to sites. Forms should be used until exist of the patient.
    • Incomplete enrollment (registration) should be avoided.
    • Completion of locator forms (both rural and urban information alongside a working telephone contact number). As many as 23% of the patients have no locator information, thus making the patient untraceable.
    • A complete channel from obtaining an AMPATH ID number to the final clinician diagnosis of the patient. As many as 3% of the LTFU patients have no charts, or the chart is untraceable. More so-even if the chart is present, it has no contents and the shell has no leading information.
  • The Death Reporting SOP needs to be re-communicated to all clinicians in all facilities. Sufficient copies of the Death Reporting Form need to be available to all sites and at all times. As many as 2% of patients were actually noted as dead in their charts yet were recorded as LTFU.
    • Protocol on death should include reporting ALL deaths to ANY AMPATH staff, not just those reported to Outreach (staff).
  • Collecting medication directly from the pharmacy with no records being filled and the information not being captured elsewhere should be avoided if the patient is enrolled in AMPATH. This is in particular reference to no encounter forms being filled. This results in patients being active on the pharmacy wing, but they are LTFU on the AMRS.
  • Privately seen patients who have only been registered at the very beginning in the AMRS system and have not come back. These patients are actually attending to services offered by AMPATH, but they cannot be accounted for. They are a drain to the resources of AMPATH as they cannot be factored into budgetary allocations.
  • Transfer In-Transfer A Out Standard Operating Procedure must be developed, including official Transfer In/Out forms should be developed and need to be filled for BOTH transfers out of AMPATH as well as to clinics within the AMPATH network. As many as 8% of LTFU patients in the AMRS have “officially” transferred according to their chart notes, but they have never been captured as such on the AMRS. More so, on a larger scale, as many as 31% of patients self transferred to a new site, were never discharged from the original clinic, had already returned to care after being lost and had evidence in their chart of an official transfer to another site.
  • AMPATH needs to adhere to patients within catchment areas. As many as 71% of patients are from in-catchment leaving 29% of patients from out-of-catchment areas who are extremely difficult to locate. They may not be defaulting at a higher rate, but when they do default, then they are the most difficult to find and bring back to care.
    Return To Clinic (RTC) dates needs to be recorded even when patients are referred to the wards / hospital.

Database Management Recommendations:

  • Database records need to be improved for:
    • A true reflection of the status of patients should be captured by the Records Department.
      • As many as 2% of patients were actually recorded as dead on their charts, but this information was not recorded on the AMRS.
      • 1% of patients, (mostly children) were discontinued from care because they were HIV negative but this was not captured.
      • 3% were never LTFU.
      • 3% had no charts at all.
  • In total, approximately 10% of the patients were wrongly captured on the AMRS . This is excluding 8% of patients who transferred out but was not captured on the AMRS.
  • It is recommended that forms to follow up children until they are 5 years old be developed.
  • RTC dates should not be over-ridden by “no date” or “upon return from ward” even when patients are referred to the wards / hospital. This is the one date that AMRS cannot do without as it records the patient’s return date for clinician’s visit.
  • Accurate appointment lists printed weekly must be properly adhered to. Subsequent to that, the weekly diary list must be re-implemented in order to have a backup system of checking those who are scheduled to come to clinics, but have not come on the appointment day. The current appointment sheets have on numerous inspections had additions of patients seen at the clinics but their appointments were “not expected” despite their patients cards reflecting that date. In this case, the Outreach Department has no way of recording patients for LTFU tracking if they do not attend.
  • There is a need to capture special cohort locator information e.g. Those whom AMPATH offers care for in prisons, truck drivers, and those who are on transit. Once these people leave their temporary places of stay, they will automatically be lost. It is possible that these groups could be given care, but not generated on LTFU lists.

Recording Keeping Recommendations:

  • It is generally recommended that the staff involved with records AT EACH site do a complete clean up of the filing rooms and to bring these workspaces to a standard that reflects a clean, neat, and safe working environment. This should be reflected by the absence of mislabeled or misfiled charts, no loose encounter forms on tables, and no loose laboratory results as the bare minimum. The filing rooms should reflect the highest organizational outlay with properly labeled shelves.
  • Because of misfiling, it is recommended that each site do a complete refilling of their patient charts. This will involve putting back charts from dead patients, those transferred, and those who have been put in boxes and ready to be discarded. This will centralize the charts and thus minimize those lost.
  • Patients must never handle their own charts.
  • Finally, the results obtained in this evaluation will most likely inform the future management of LTFU including the regular analysis on mortality, retention into care, and resource allocation. It is recommended that more resources be utilized in actively finding patients before they are completely disengaged from care.
  • It is also expected that this process of intensively finding those LTFU will be fed back to CHWs and outreach staff alongside data clerks at individual sites. This will form part of the supervisory feedback and monitoring of implemented changes.
  • Now that baseline information has been collected, this should be tied to the above feedback process. It is thus recommended that periodic (possibly every 2 years) Outreach initiated evaluation similar to this one be done.