|
Caroline Bowyer is senior pharmacist
elderly care and NSF at Southampton City Primary Care Trust
e-mail
caroline.bowyer@scpct.nhs.uk
Nina Barnett is
pharmacy adviser for older people, London SE and Eastern Specialist
Pharmacy
|
The HOMER trial (“Does home medication review keep older people
out of hospital?”) published earlier this year at www.bmj.com and
reported in the PJ (22
January, p71) had some unwelcome results for pharmacists: the intervention
(pharmacist) group increased hospital readmissions by
30 per cent and GP home visits by 43 per cent. These results were collected
from a group of 872 elderly patients discharged from hospital following
an emergency admission. The participating pharmacists also appeared to
decrease the patient’s quality of life as well as non-statistically
decreasing the risk of death.
The BMJ article concluded that this study showed “home based medication
reviews by pharmacists may increase hospital admissions” and “more
effective forms of medication review need to be established considering
patients’ quality of life and effects on both hospital and general
practice, as well as prescribing outcomes”.
Is this the end?
Is this the end of pharmacist medication reviews as we know them? Or
is this an opportunity to take stock, learn and move forward?
On closer examination of the trial, it is clear that the pharmacists
participating in this trial were charged with a difficult task. The pharmacists
saw elderly patients (average age 85.5 years), recently discharged after
an emergency admission, with only a discharge summary to hand. Discharge
summaries do not always reflect a patient’s complete medication
list and with no access to medical records, the pharmacists could not
verify the discharge medication list for any of the patients. The authors
maintain that having no access to medical records merely mimics a “medicine
use review”, an advanced service under the new pharmacy contract.
However, community pharmacists undertaking these reviews would have both
access to the PMR at the pharmacy and quite possibly personal knowledge
of the patient to support them.
It is interesting to note that 67 of the 429 (16 per cent) patients in
the intervention group did not receive a medication review. In the analysis,
which was intention-to-treat, these 67 were counted as intervention patients
and if they were admitted to hospital this would have been recorded as
being due to a (non-existent) medication review. Another interpretation
However, we would like to suggest another interpretation of what the
HOMER authors concede is a counter-intuitive conclusion from their
study. It is possible that this trial showed that pharmacists unmasked
the clinical needs of patients which were then not met, increasing
hospital admission. The pharmacists visited the patients an average
of seven days after discharge working in isolation, rather than part
of a multidisciplinary team. There appeared to be no system in this
trial which allowed the pharmacists to refer the patient to other non-medical
health care professionals, who may have helped prevent admissions.
We know that some admissions have a medication-related component, but
there are many which are due to complex social and medical issues. This
study assumes that all elderly admissions are iatrogenic or have an iatrogenic
component which can be influenced by a pharmacist intervening. This patient
group would have been especially vulnerable given their recent emergency
admission and that 15 of the intervention group patients had three or
more admissions in six months, in addition to the index
admission.
These “revolving door” patients must surely have complex
histories and would be worthy of entry in an Evercare or Kaiser Permanente
managed care type project. It is unlikely that medication review alone
could be shown to affect the outcomes in these patients. Perhaps hospital
admission is a poor marker for iatrogenic morbidity and mortality. It
is possible that the process is so complex as to preclude using this
study method to look at the effect of pharmacist intervention in isolation.
In the past few years we have seen a swift increase in rapid response
teams that prevent admissions in vulnerable patients. These patients
would never be seen in this study. Emergency hospital admissions for
older people tend to occur at times of crisis, for instance when the
patient’s main carer is suddenly unavailable or the patient suffers
a serious medical or surgical event. These common reasons for admission
are unlikely to be directly related to medication review and underline
the need for publication of reasons for admission in this study. A recent
BMJ letter1 suggests that this
cohort of patients was not representative of the general population of
elderly in the community and that there
are no details of how pharmacist interventions in this study were designed
to address issues of iatrogenic hospital admission or, indeed, any other
reasons for
admission.
There is robust evidence in the UK supporting clinical medication reviews2 in
a GP surgery. This study demonstrated significant improvements in terms
of medicines management and number of drugs prescribed, without
an increase in GP time. Further work is required to assess the effect
of this work on hospital admission rates.
The published data from the HOMER study leave us with more questions
than answers. We need to see more details of the reasons for admission
for all HOMER patients. The inclusion of an elderly care physician from
the acute trust would help in the assessment of admissions. It would
also be interesting to see the data represented after the exclusion of
the 67 patients who did not receive a medication review to ensure no
data bias. We hope that this information can be retrieved retrospectively
from hospital notes. A cautionary note
What the HOMER trial has done for pharmacists is raise an important
cautionary note about medication reviews, the competency and clinical
knowledge
required to do them, the interventions suggested and the way they may
impact on frail elderly patients. A recent quote in the PJ from the
Medicines Partnership that “medication use review is a fantastic
opportunity for GPs
to make use of pharmacists to carry out face-to-face
reviews on their behalf” (PJ, 26 March, p351) may not
be seen as such by GPs reading the HOMER trial. We look forward to
the publication
of further analysis of the HOMER data to enhance our understanding
of this complex area.
References
1. Hay JW. Pharmacist medication review study design concerns. BMJ
2005;330:E347.
2. Zermansky AG, Petty DR, Raynor DK, Freemantle N, Vail A, Lowe CJ.
Randomised controlled trial of clinical medication review by a pharmacist
of elderly patients
receiving repeat prescriptions in general practice. BMJ 2001;323:1340. |
The
Pharmaceutical Journal