Advances in the Canine Cranial Cruciate Ligament, 2nd edition

318 Surgical Treatment

adverse event if they were in the control group.
Similarly, Absolute Risk Increase(ARI) can be
calculated.
To calculate NNT and NNH, it is neces-
sary to choose a specific outcome measure at a
specific time point. For example, if success at
1 year after surgery is defined as at least a 25%
improvement from a validated owner survey
and no residual lameness via visual inspection
12 months after surgery, failure will be defined
as the need for a second surgery. For the control
group it can be conservatively estimated that
with non-surgical management of large-breed,
obese dogs with cruciate ligament rupture (CR),
50% of dogs will have a successful outcome
and 0% of dogs would require a second surgery
(Wuchereret al. 2013). For tibial plateau leveling
osteotomy (TPLO) it is estimated that 90% have
a successful outcome (owner reports improve-
ment plus no visual lameness) and 10% have
failure (e.g., second surgery needed to remove
bone plate). The present authors used 10%
based on unpublished data from their hospital,
where it was found that over a 5-year period
9% of patients that had a TPLO plate applied
underwent a second surgery where the TPLO
plate was removed. For comparison, only 80%
of dogs that have an extracapsular suture placed
have a successful outcome, and 5% require a
second surgery. Using these estimates of suc-
cess and failure, it can be learned that both pro-
cedures provide a good treatment advantage
over non-surgical management and TPLO pro-
vides greater benefits (Table 39.1). However, if
the relationship between NNT and NNH (a low
number is good) is reviewed, a conclusion that
TPLO is much better than a suture surgery after
examining only NNT is an incomplete interpre-
tation, as the risks of TPLO are greater than for
a suture surgery.

Owner questionnaires

Establishing the owner’s perspective on out-

come remains important since they spend

the most time with their pet. Historically,

manuscripts have documented that owners

reported that their pet had a ‘good or excel-

lent’ outcome 88.5–93% of the time after surgery

(Hoffmannet al. 2006; Corr & Brown 2007;

Stein & Schmoekel 2008). These findings, how-

ever, must be interpreted with great caution

because of differences in the questions asked,

the method of questionnaire delivery (mail,

telephone, or personnel interview) and other

biases that significantly influence owner report-

ing. It is also important to note the duration

of follow-up after surgery, since it has been

reported that outcomes worsen with increased

time after surgery (Inneset al. 2000). To best

interpret and compare studies, the consistent

use of a validated owner questionnaire should

be considered. Several validated owner ques-

tionnaires addressing patient pain, function,

quality of life, and/or activity have been cre-

ated. The Canine Brief Pain Inventory (CBPI)

for dogs with osteoarthritis (OA) provides rea-

sonable evidence that owner questionnaires can

be effectively used in clinical research (Brown

et al. 2008). The Liverpool Osteoarthritis in Dogs

(LOAD) instrument has been critically evalu-

ated and successfully used in several clinical

investigations (Waltonet al. 2013; Lascelleset al.

2015). The Client-Specific Outcome Measures

(CSOM) has also been evaluated and used in

clinical research (Riallandet al. 2012).

While each of these questionnaires will

effectively standardize the questions and allow

for some comparison across studies, results in

studies with small to modest patient enrollment

numbers should be interpreted with caution

Table 39.1 An example of how NNT and NNH can be used for a balanced approach to compare different interventions.

Success

rate (%)

NNT

1/Tx-control

(control=50%)

Major complication

(2nd surgery) (%)

NNH

1/%

complication NNT/NNH

TPLO 90 2.5 10 10 0.25

Extracapsular suture

stabilization

80 3.33 5 20 0.167

NNT, Number Needed to Treat; NNH, Number Needed to Harm; TPLO, tibial plateau leveling osteotomy.