To The Editor:
We are writing in response to "Commentary. The Fallacy
of Short-Term Outcomes Analysis in Pediatric Orthopaedics" (81-A:
1499-1500, Oct. 1999), by Winter. As co-investigators in an OREF-funded study
examining the utility of quality-of-life questionnaires in the evaluation
of patients with pediatric orthopaedic problems, we share Dr. Winter’s interest
in continuing to improve our understanding of the outcomes in children
after clinical intervention. However, with all due respect to Dr.
Winter’s experience and perspective, we strongly disagree with
his call to "abandon efforts to use outcomes instruments
as they now exist" in this area.
Dr. Winter is quite correct in highlighting the intrinsic difficulty
in the assessment of outcomes in children. Children continue to
develop and mature despite illness and during treatment, creating
a "moving target" when assessing functional and psychosocial
abilities. While we are generally interested in the patients’ perceptions
of their own health, it is obviously difficult to obtain such information
directly from our younger patients, forcing us to rely on the parents’ assessment.
Children often have known no other health state and may have adapted
to disability, yet we still would hope to offer even the most chronically
involved children a better quality of life if we were able. Moreover, many
conditions in our field are rare and do not lend themselves well
to prospective study. Finally, given the fact that some procedures
in pediatric orthopaedics are prophylactic in nature, there is a question
as to when we should assess the effect of our surgery.
In focusing on this last issue, Dr. Winter ignores numerous areas
of pediatric orthopaedics in which intervention is aimed at improving
health status in the short term (for example, trauma, limb deficiency,
soft-tissue surgery in cerebral palsy, and even oncology), areas
in which "short-term outcomes analysis" informs
us about what we have done. Dr. Winter also points out that, in
scoliosis, we operate to prevent the morbidity associated with curve
progression. However, current evidence suggests that only a subset
of children who are candidates for surgery will develop significant cardiopulmonary
problems as adults if untreated; clearly, we base the decision to
operate on other factors as well. Numerous authors have documented
the effect that scoliosis has on cosmetic appearance, psychosocial
function, and quality of life, and we should not only recognize
these as factors in the decision-making process but should strive
to better understand them. Newly available health-status measures
can quantify the effect that scoliosis has on health status, broadly
defined to include issues related to psychosocial function and quality
of life, and we should add such measures to our armamentarium. As
such, information regarding psychosocial function and self-image
obtained two years after surgery does not threaten to replace information
regarding pulmonary function obtained forty years after surgery,
but, rather, complements it.
Despite the intrinsic difficulties in assessing outcomes in children,
it is comforting to know that we pediatric orthopaedic surgeons
are not alone in having to deal with these issues. Our colleagues
in general pediatrics and in various pediatric subspecialties are
also constantly confronted with the same issues. Furthermore, a
growing cadre of professionals trained in clinical research methodologies
(including quality-of-life assessment) has emerged to help us, as
busy clinicians, to apply all tools at our disposal to assess more
rigorously the effects we have on the children we treat. The recent publication
of the AAOS/POSNA pediatric orthopaedic instrument highlights
the efforts of our leadership to do just that1.
We have developed powerful and effective treatments for pediatric
orthopaedic problems. However, our ability to document the effects
of clinical intervention has not evolved at an equal pace. We must
continue to refine all methods at our disposal to better understand
the effects of what we do. Quality-of-life measures provide one
more modality, one more perspective, on outcomes that should be
considered. Perhaps we should add to Dr. Winter’s description
of the pendulum process that after "overenthusiasm" and "rebellion" should
come the appropriate incorporation of the technique into our clinical
armamentarium.
R.B. Winter replies:
Drs. Vitale and Roye have failed to read carefully what
I wrote. At no time did I condemn outcomes analysis in its entirety
when discussing children’s orthopaedics.
My comments were specifically focused on the issue of preventive
treatment, not on the child with already measurable disability.
A good example is a one-year-old child with congenital scoliosis
due to a unilateral unsegmented bar. The appropriate treatment is
immediate arthrodesis to prevent curve progression and subsequent
disability (early death due to cor pulmonale).
Since this one-year-old child has no measurable disability
and is unable to fill out a questionnaire, and since there is no
possible questionnaire that the parent can fill out, it is totally
impossible to apply the concept of outcomes analysis to this situation other
than to see whether or not this patient dies early or has a long
and useful life span.
It is perfectly appropriate for pediatric orthopaedists to develop
outcomes analysis capability when (a) there is an already existing
disability for which treatment is being rendered and (b) there is
a valid method for measuring the quantity and quality of that disability
both before and after the treatment.
Adolescent idiopathic scoliosis is a "slippery" arena
in which to apply outcomes analysis since there is a mixture of
preventive measures and immediate treatments for correction of existing malfunction
to be considered. Outcomes measurement tools that address only the
cosmetic and "psychosomatic" elements over a short
term (two-year follow-up) ignore the more important prevention of pulmonary
malfunction (not cor pulmonale) and arthritic problems during adult
life.
I do not mind if they look at cosmetic improvement as long as
they also state in their report that their analysis does not include
the more important half of the equation.
In conclusion, I agree with the statement that the pendulum should
rest in the center, avoiding both underenthusiasm and overenthusiasm.
My position is that we have not yet reached that centerpoint, and efforts
to "make official" that which has not been adequately
thought out are dangerous.