Method and results

1
Follow-up or discharge? A new patient outcome
analysis
Anna Litwic, Andrew Bamji, Queen Mary's
Hospital, Sidcup, Kent DA14 6LT and Queen
Alexandra Hospital, Portsmouth PO6 3LY
Method and results Details of 400 consecutive new
patients were entered onto the spreadsheet, and
analysed by diagnostic group and follow-up
pattern. The results were compared with
historical data (Bamji, Dieppe, Haslock and
Shipley) and against contemporary audit data from
a neighbouring hospital (Yanni, personal
communication). Numbers. ia
inflammatory arthritis (141) sp spinal
condition (neck/back pain) (65) bo bone
disorder (9) cr crystal arthritis (5) sh
shoulder disorder (28) oa osteoarthritis (any
site) (51) st soft tissue injury excl shoulder
(81) ot non-rheum diagnosis(7) qd diagnosis
uncertain (9) cp chronic pain syndrome (4)
total 400 As expected, follow-up was heavily
biased towards RA and IJD. Analysis also showed
that these categories tended to earlier follow-up
than other conditions. Reasons for not booking
follow-ups varied for some conditions (e.g. back
pain) a onward referral to physiotherapy was the
primary reason while for others appointments were
suspended pending investigations or were booked
to reassess after investigation, and initial
treatment. The results confirmed that one-stop
patients would not generally be followed up
overall 53 of patients received no immediate
follow-up appointment. 85 of patients with
inflammatory conditions including RA did get a
further appointment ( a slight underestimate
because some patients had suspended
appointments pending investigation) , but only
20 of non-inflammatory conditions received one.
It thus became clear that the difference in
overall new-follow-up ratio between Queen Marys
and Lewisham could be at least partly accounted
for on the basis of a difference in casemix, as
shown below. Diagnostic category
casemix comparison between Lewisham Hospital (100
consecutive referrals) and QMH Sidcup (400
referrals). The relative proportions of IA
(likely to generate follow-up) and back pain
(one-stop) are reversed (Lewisham data courtesy
of Dr Ghada Yanni)
Introduction The costs of secondary care should
be justified by analysis of patient outcomes.
There is a perception that many patients seen in
hospital outpatient clinics are followed up
unnecessarily and for protracted periods thus
pressure is applied to rheumatology departments
to reduce follow-up numbers with the aim of
saving money. The ways in which this has been
approached are crude and often based on attempts
to align departments ratios of newfollow-up
appointments. Such policies may conflict with
disease management guidelines and it is thus
important to understand the casemix of ones
department and the likely effects of this on the
need for continuing follow-up. In addition an
analysis of new patient casemix can provide
useful information on the numbers of one-stop
appointments. It is also necessary to audit
changes in follow-up patterns if new policies are
adopted. The collection of such data allows
explanation of practice, change and differences
between units and may enable the successful
defence of traditional outpatient care. Ongoing
analysis of new patient referrals will also
detect changes that occur as the result of
outside pressures, such as the introduction of
referral management systems or the development of
alternative provision.
In 1990 we noted differences between centres in
the numbers of new patients seen with specific
diagnoses as shown on the graph
below. Comparison of SE London (SEL),
The Middlesex (L), Bristol Royal Infirmary (Br)
Middlesbrough (Mi) Hospitals (from Bamji, Dieppe,
Haslock Shipley, 1990) . NT and T refer to
teaching hospital status. IA includes RA
IJD Direct comparison of data from a single unit
(QMH) shows that there has been a significant
shift in casemix pattern. We assumed
this could be accounted for in particular by the
introduction of a physiotherapy-led back pain
triage service which significantly reduced the
numbers of spinal disorders attending the
rheumatology new patient clinic which also
explains the disparity with Lewisham Hospital,
which at the time of analysis did not run such a
service. In one year nearly 1500 patients were
seen in the QMH triage service. We calculated
that, if they had been included within the
rheumatology figures, the newfollow-up ratio
would have reduced from 14.2 to 11.7 on the
basis of expected outcome and follow-up pattern.
This underlines the dangers of uncritical
comparison of crude outpatient figures as a basis
for planning care and the need to include casemix
in any comparative exercise. However it is
apparent that the major change in percentage
terms has been the reduction in numbers of soft
tissue and shoulder problems seen. Discussion
with local GPs indicates that this is due to an
increased number of these problems being treated
in general practice. The relatively minor shift
in OA/spine new patients may also reflect the
deliberate withdrawal of low back pain patients
from the orthopaedic service those still seen in
rheumatology also include patients who have been
through the triage service once and for whom
automatic triage straight to rheumatology is
indicated according to the service protocol.

• Figures from 1990 (1) show clearly that IA
  attracts a much higher newfollow-up ratio than
  NIC and, indeed, our results show that this
  remains the case. However we decided to target
  follow-up in an attempt to reduce the number of
  IA follow-up patients (as requested by one PCT)
  using clinical criteria. These were
• Elongation of F/U interval where appropriate
• Compliance with BSR guideline of minimum 1
  review/yr
• Discharge if stable for gt 3yr unless on biologic
  agents (approx 230 IA patients)
• The graphs indicates the success of this measure
  however the overall reduction in F/U appointment
  numbers did not reach the level requested by the
  PCT largely because the area for maximum impact
  (NI) was the smallest group.

Graphs showing the change in follow-up patterns
as the result of explicit decision-making on the
need for continuing care for RA (top) other IJD
(middle) and NI (bottom) between October 2006 and
March 2008. The X-axis indicates the time to next
appointment the Y-axis indicates numbers of
patients. The former two groups show a shift to
the right implying that patients with IA are
being seen less frequently. The latter shows a
substantial drop in the total of NI patients
remaining in regular follow-up but represents
only 5.6 of the total follow-up group.
Data collection The rheumatology department at
Queen Marys Sidcup has developed a simple Excel
spreadsheet for recording patients on long-term
follow-up, with data on patient age, PCT origin,
diagnosis, treatment, length of follow-up and
time to next appointment. Onto this have been
added separate analysis sheets for audit
projects in addition to a breakdown of casemix
RA, other inflammatory joint disease (IJD) and
non-inflammatory conditions (NIC) that are in
long-term follow-up we collect data on new
referrals, discharges (by cause) and patients on
biologic agents. The spreadsheet also allows
automatic calculation of the Disease Activity
Score (DAS). Thus at any time we can take a
snapshot of clinic activity, track changes in
follow-up pattern and check new patient casemix.
The database template is available through the
departments website at www.sidcuprheum.org.uk/bsr
actionplan.html We expected that arranged
follow-up from first consultation would be skewed
in favour of RA and IJD but were not clear why
our newfollow-up ratio (14.2) was twice that of
a neighbouring unit (Lewisham University
Hospital, 12.1) which we assumed would have a
similar casemix. We also wished to confirm that
patients with NIC, soft tissue lesions, shoulder
pain and spinal conditions were indeed treated in
a one-stop fashion rather than simply relying
on an assumption that this was the case - though
this was likely from the numbers in follow-up
which were RA 529 IJD (including
seronegative arthritis, SLE, vasculitis,JIA) 473 N
IC 56 The numbers in follow-up are not
necessarily directly related to appointment
numbers patients in an unstable state will have
more frequent appointments and the offer of
emergency open-access will also increase overall
numbers.
Conclusion A historical comparison of casemix
indicates that (1) it is possible to change
patterns of practice (2) a substantial proportion
of patients seen does not require any follow-up
(3) use of crude newfollow-up ratios to compare
hospital performance will mislead if it fails to
account for casemix differences but (4) any shift
of caseload away from a traditional mix of IJD
and non-inflammatory musculoskeletal conditions
will change follow-up patterns and may
concentrate care on more complex and thus more
expensive conditions.
References Bamji A, Dieppe P, Haslock I, Shipley
M. What do Rheumatologists do? A Pilot Audit
Study. Br J Rheumatol 1990, 24,
295-8 Correspondence andrew.bamji_at_qms.nhs.uk