EDITOR – In their editorial on the two week rule for cancer
referrals, Jones et al discuss the fact that steps to meet the target for
“urgent” referrals have led to a doubling of waiting time for “routine”
cases [1]. This is exactly the result we would expect from our own
calculations modelling waiting times with a Monte-Carlo model based on
Poisson fluctuations in demand [2].
To sustain a waiting time below two weeks, capacity needs to exceed
mean demand by approximately two patients per week, for a wide range of
values of mean demand. This applies to any appointment, including those
for diagnostic and staging procedures, as well as for treatment. Applying
this excess capacity to a sub-group of “urgent” patients is inherently
less efficient than applying it to reduce the waits for all patients. If
a fast-track for “urgent” referrals is created by transferring resources
from “routine” patients, this is likely to lead to demand exceeding
capacity for “routine” patients, and hence to ever-increasing waiting
times.
These calculations assume that variations are due to random
fluctuations in a constant demand; this has been shown to be an effective
model for oncology referrals [2]. There is some evidence from dermatology
[3] that reductions in waiting times actually lead to increase in demand,
negating the benefit of increased resources. It is possible that this
effect may also apply in referrals for suspected cancer, in which case the
extra capacity required will exceed our calculations. This effect should
not apply to the number of cases of diagnosed cancer, which will be
limited by the incidence of the disease. This suggests that resources may
be more effectively targeted on reducing the waiting times from diagnosis
to treatment, than on reducing the time from GP referral to diagnosis.
Rapid Response:
capacity needs to exceed mean demand
EDITOR – In their editorial on the two week rule for cancer
referrals, Jones et al discuss the fact that steps to meet the target for
“urgent” referrals have led to a doubling of waiting time for “routine”
cases [1]. This is exactly the result we would expect from our own
calculations modelling waiting times with a Monte-Carlo model based on
Poisson fluctuations in demand [2].
To sustain a waiting time below two weeks, capacity needs to exceed
mean demand by approximately two patients per week, for a wide range of
values of mean demand. This applies to any appointment, including those
for diagnostic and staging procedures, as well as for treatment. Applying
this excess capacity to a sub-group of “urgent” patients is inherently
less efficient than applying it to reduce the waits for all patients. If
a fast-track for “urgent” referrals is created by transferring resources
from “routine” patients, this is likely to lead to demand exceeding
capacity for “routine” patients, and hence to ever-increasing waiting
times.
These calculations assume that variations are due to random
fluctuations in a constant demand; this has been shown to be an effective
model for oncology referrals [2]. There is some evidence from dermatology
[3] that reductions in waiting times actually lead to increase in demand,
negating the benefit of increased resources. It is possible that this
effect may also apply in referrals for suspected cancer, in which case the
extra capacity required will exceed our calculations. This effect should
not apply to the number of cases of diagnosed cancer, which will be
limited by the incidence of the disease. This suggests that resources may
be more effectively targeted on reducing the waiting times from diagnosis
to treatment, than on reducing the time from GP referral to diagnosis.
Simon Thomas
physicist
simon.thomas@addenbrookes.nhs.uk
Neil Burnet
consultant oncologist
neilb@argonet.co.uk
Addenbrooke’s NHS Trust,
Cambridge CB2 2QQ
1. Jones R, Rubin G, Hungin P. Is the two week rule for cancer
referrals working? BMJ 2001;322:1555-6
2. Thomas SJ, Williams MV, Burnet NG, Baker CR. How Much Surplus Capacity
is Required to Maintain Low Waiting Times? Clinical Oncology 2001;13:24-
28
3. Smethurst DP, Williams HC. Are hospital waiting lists self-regulating?
Nature 2001;410:652-3
Competing interests: No competing interests