No. Some studies suggested that while decentralisation may
increase the reach of the programme, the efficiency may not increase or the
quality of care can suffer. One problem is that there is much less
infrastructure and diagnostic equipment which can mean delayed diagnosis for
those attending remote primary care clinics. For instance, in one very large
study of 663 health facilities managing over 900,000 HIV-positive patients
across nine countries in sub-Saharan Africa
– those who attended large tertiary
hospitals had better access to the more advanced TB diagnostics.
Since it doesn’t always achieve the desired effects, it is
very important that clinic teams and programmes know how integration of service
delivery is working for them. This means research and maintaining good records
at the health facility.
Sometimes there isn’t
good enough data to be able to know for certain. Because of poor data reporting it is often
difficult to gauge the success of a programme. For example, there was one
report at the IAS meeting about an effort to introduce intensified TB case
finding into ART clinics in Lusaka,
Zambia. To
facilitate the service delivery,
a worksheet was produced to improve the quality and consistency of symptom
screening (and link it to the diagnostic work-up). They provided nurses and
clinical officers with intensive training and mentorship on TB/HIV management
over three to four years.
Documentation
in the patients' files suggested that staff were asking the patients about symptoms
of TB. But the healthcare workers had refused to fill out the TB screening
worksheet because it was seen as extra work. Yet without filling out the forms, there was no way of telling whether
patients who screened positive had been sent for diagnostic work-up, and what
their outcomes were.
Take home message: when a programme integrates any new
services it must make certain that the data are being recorded.
When new
services are introduced into a facility and programme, it is critically
important to fill out the associated registers
– otherwise, it can be very
difficult to establish how implementation is going, whether the services are
being performed consistently, and whether they are cost-effective. When staff
do not record the data, they may just wind up dooming themselves to keep
providing an unwanted or non-effective service, repeating actions that are a
waste of time or worse, a distraction from more useful services that they could
be providing. Having an answer to these questions is in the best interest of
patients and of the people providing these services! Without reliable data, it will also be impossible for clinic teams
to engage in quality improvement exercises.
In other cases, the
data are contradictory, either because it takes time at first for provision of
a new service to reach its stride, or there is a right way or wrong way to
provide a new service or decentralise.Case
in point: findings are mixed on NiMART itself, and the decentralisation of HIV
care to the primary health centre.
Virtually everyone agrees with giving nurses at the primary
healthcare level the primary duty of starting and managing ART in people living
with HIV.
But some people think the shift over to NiMART and
decentralisation should be faster, while others advise more caution, warning
that gaps and problems emerge when the transition to poorly supported nurses at
poorly equipped primary health centres happens too quickly.
There were studies presented in support of each camp.
In Malawi,
there are very few doctors, so ART services are delivered by clinical officers,
who manage the most complicated patients, and nurses who manage the rest. An
observational cohort study by MSF in Malawi found much worse outcomes
among the patients managed by clinical officers, than those cared for by the
nurses. But since the clinical officers
took care of patients who were more ill, what is most important is that the
nurses did a good job managing PLHIV.
A strong partnership with trained patient networks improves linkages and keeps people in care
In South
Africa, the STRETCH study found a nurse-led
ART service was just as effective as a doctor-led service, and even seemed to
provide better quality of care in some regards. But contrary to expectations,
nurse-managed care did not seem to increase access to ART. But in this study, NiMART hadn’t been fully
introduced to all centres in the study and nurse confidence in prescribing ART
should improve significantly with more experience. Performance may get better
with time.
But there might be factors affecting access and retention
other than whether a nurse or a doctor is initiating ART. In Mozambique,
integrating HIV services into primary healthcare centres run by clinical officers dramatically
expanded coverage compared to when ART was only available from doctor-based ART
sites. But patients attending the primary healthcare sites were also much more likely to drop
out of care. When the researchers asked
why, the patients told them that the ART sites had dedicated pharmacies for
them, but the pharmacies at the integrated clinics were of poorer quality, and importantly,
offered the PLHIV no privacy. In another
study in South Africa,
patients refused to be down-referred to their local primary healthcare clinic
because they felt that it was under-resourced and wouldn’t provide the various support
services available at the larger health facilities that they were used to
attending.
Ongoing discussions to re-engineer primary health care must
address these problems convincingly. New point-of-care lab tests may enable
nurses to address some of the patient’s concerns, if they are rolled out to the
primary health clinic.
Other primary health clinics take the specimens onsite and
send out for timely diagnosis. One study described the successful
implementation of integrated TB and HIV services decentralised to a very busy
primary healthcare clinic in Johannesburg. Even though the clinic has support from a
doctor on three days a week, it sees more than 8500 patients per month, with
4200 people currently on ART being managed there. In addition to general HIV services, the clinic
provides TB screening and treatment, pre- and post-natal care, general/chronic care,
paediatric care, social services and a pharmacy. Blood for CD4 cell testing is
drawn at the clinic, and sent off-site. Likewise, TB diagnostic services, such
as chest X-ray, microscopy, and culture have to be performed off site.
Despite being a busy place, the clinic achieved a very
high uptake of HIV testing among people diagnosed with TB attending the clinic
for treatment, and patients also eligible for ART started treatment promptly.
The clinic
has community-based tracking procedures in place to make certain that TB
patients get their diagnosis in a timely fashion, so that they can start
treatment as quickly as possible. Another study,
on an integrated one-stop shop TB/HIV clinic in Uganda, explained how an active
tracing system was established and employed whenever patients missed
appointments. It worked extremely well.
In fact, what seemed to often determine the success of
integrating services or decentralisation
to the primary health clinic, was the community, particularly the engagement of
people living with HIV in the service. A strong partnership with
trained patient networks improves linkage and keeps people in care. Several studies reported how
the training and employing expert patients dramatically improve outcomes. In
the study from Northern Uganda, the researcher
concluded that having community-based peer supporters who would help follow up
and provided community sensitisation of women for the antenatal services was
critical.
This
idea is not new. The most common model for community engagement is based on the
work of Partners In Health with accompagnateurs, community members and expert
patients who accompanied referrals and supported patient in Haiti and Rwanda.
Partners
Health has developed an accompagnateur training guide in several languages for
programmes and clinic teams who would like to train and employ their own expert
patients.