In November 2001, after obtaining medical ethical board approval, until January 2005, all 87 infants referred to Starship Children’s Health for clubfoot were offered the options for nonsurgical (Ponseti) or surgical management of clubfoot. Families were given the option of randomization or choice of treatment; only nine of these patients’ families agreed to randomization. Selection criteria for this review required at least 2 years of clinical followup from the initial casting to be included. In addition, only patients with idiopathic talipes equinovarus were included. Fifty-five patients, with an average age of 20 days (range, 10–34 days) at presentation, with 86 feet met these inclusion criteria. Twenty-six patients (40 feet) were in the Ponseti group and 29 patients (46 feet) were in the below-knee casting and surgery group.
At the time of referral to Starship Children’s Health for clubfoot, patients’ families met with a dedicated clubfoot nurse coordinator (JED), and treatment options, including either the Ponseti method or below-knee casting and surgical correction, were discussed. The risks and benefits of each method were explained in detail. The family was given informational handouts describing each clinical pathway and was directed to appropriate web sites. Once a clinical pathway was chosen, the patients choosing the Ponseti method were placed under the care of one senior author (HAC) and those in the surgical arm were placed under the care of the other senior author (SJW).
Demographic comparisons between the two prospective cohorts showed no difference in terms of gender, ethnicity, family history, bilaterality, age at first casting, average initial Pirani score (5.0), or years of followup (Table ). We prospectively collected clinical data at each clinic visit using a templated data sheet (Appendix 1). At the time of presentation, the clubfoot deformities were graded using the validated 6-point scale of Pirani et al. [9
]. This was typically performed by either one of the senior authors or the clubfoot nurse specialist. Other information routinely obtained included ethnicity and family history. Ethnicity was classified as Polynesian or non-Polynesian.
Demographic comparison between cohorts
In the Ponseti group, feet were treated at weekly intervals as described by Ponseti [19
]. This involved above-knee casting followed by percutaneous tenotomy of the Achilles tendon when indicated and final casting, which remained in place for 3 weeks. All patients in the Ponseti cohort who were compliant with weekly casting required tenotomy. Indications for tenotomy were an inability to dorsiflex the ankle at least 20° and any case the senior author believed the tendo-Achilles to be tight. One of the senior authors (HAC) managed this cohort as he had been directly taught by spending 1 year with Dr Ponseti. At the completion of cast treatment, all babies were placed in a foot abduction brace. The open-toed, high-top shoes (MJ Markell Shoe Co, Yonkers, NY) were fitted by the clubfoot nurse coordinator and attached at shoulder width to the Denis Browne bar. Parents were emphatically instructed to ensure full-time brace wear for 3 months followed by night and naptime wear until the age of 2 years [10
]. Any problems with casts, brace wear, and brace compliance were noted. As in our previous report, compliance was defined as full-time brace wear for 3 months followed by at least 9 months of nighttime and naptime use [10
In the surgical group, feet were treated with below-knee plaster casts applied over Elastoplast tape (BSN Medical, Auckland, New Zealand). These casts were applied with upward pressure directed under the cuboid. This was believed to help evert the foot and gradually correct the equinus deformity. These below-knee plasters were changed at weekly and biweekly intervals until approximately 6 months of age. When indicated, surgery was then scheduled. The majority of these patients required a posterior or posterior medial release and subsequent plaster changes. The surgical procedures were carried out through a Cincinnati incision [5
]. If the patient had isolated hindfoot equinus at the end of casting, a posterior release involving a Z-lengthening of the tendo-Achilles, tibiotalar joint release, and posterior subtalar joint release was performed. Patients with both forefoot and hindfoot involvement underwent a full posterior medial release. The subtalar interosseous ligament and deep deltoid ligament were left intact. Patients typically had one subsequent plaster change under anesthesia for a posterior release and two for a posterior medial release. After final plaster removal, patients were followed clinically but required no further brace wear. However, patients who corrected after casting alone (two of 29 patients) were prescribed open-toed, high-top shoes attached at shoulder width to the Denis Browne bar. Patients were then followed at regular intervals as deemed appropriate by the treating surgeon.
Clinical records of all patients in these groups were reviewed. We noted all surgical interventions and complications associated with these deformities. Initial consultation, clinic visits, and all procedures (primary, revision, complications) until last followup were recorded. The average age at followup in the Ponseti group was 3.5 years (range, 2.16–5.5 years) and in the surgical group 3.8 years (range, 2.2–5.7 years). Nonmonetary measures, including operating room time, days in the hospital, number of anesthetics received, and medication required, were reviewed. Using clinic and hospital billing information, the cost of care per foot was then determined. These costs were determined based on the New Zealand public health hospital system. All dollar amounts were then adjusted to US dollars (using a conversion rate of NZ $0.75 to US $1.00). Using this clinical data and corresponding CPT codes, the cost of care for these patients as if they had been treated in the United States was calculated. Variables used in determining clinic costs included consultation, casting, foot abduction braces (Markell shoes and Denis Browne Bar), and in-office procedures. Variables used in determining hospital costs included operative time, surgeon and/or anesthetic fees, perioperative care, and number of days in the hospital. Medications are included in the operative and hospital costs in the New Zealand health system and were not billed separately; therefore, these were not included in the cost of care in the US system. Similarly, cost of time in the postanesthesia care unit is combined with day stay or hospital admission costs and not billed separately in the New Zealand system. Therefore, these charges were not entered in the calculations for care within the US healthcare system. For patients awaiting procedures, estimates of operating room time and days in the hospital were based on other patients within the same treatment group undergoing an identical or similar procedure. To determine the end point of clinic visits in the New Zealand system, we used the discontinuation of abduction orthosis in the Ponseti group and the first postoperative visit in those children requiring surgery. This was believed to indicate the end of initial “active treatment” and remove any subjective differences in followup between cohorts. However, any patient found to have a recurrence or complication in either cohort requiring further treatment had these hospital costs included in the final calculations (See Secondary Surgical Costs in Tables and ). For determinations involving the US system, one visit outside of the 90-day global period was assumed as the end point for routine care (for those feet without recurrences) in those patients requiring surgical management.
New Zealand cost comparisons converted to US dollars
Comparisons to US cost (cost of treating these cohorts in the United States)
Demographic nominal variables (gender, ethnicity, unilateral versus bilateral, and family history) were compared in the two groups by a two-tailed Fisher’s exact test, continuous values (age at first casting, initial Pirani score, and age at final followup) by the Student’s t test. We determined differences in nonmonetary resources (number of clinic visits, hospital days, anesthetics, operative time, antibiotic doses, and pain medication doses) between the Ponseti and surgical cohorts using the Student’s t test. To more accurately assess monetary differences between cohorts, patients within each cohort were further subdivided into unilateral and bilateral cases as well as those with and without recurrences. We then determined monetary differences between each subgroup treated with either the Ponseti or surgical method using the Student’s t test. Separate analyses were performed for care given in the New Zealand socialized healthcare system and for the estimated cost of care in the US healthcare system. Finally, a paired Student’s t test was used to compare differences in total cost to manage these feet by the Ponseti method or surgery in each of the healthcare systems.