Resection of Tarsal Coalition in 27 Children with 2 Years Follow-Up – Patient-Reported Outcomes Using the Validated Oxford Ankle Foot Questionnaire

Abstract

Background

Patient Reported Outcome Measures (PROM) after resection of tarsal coalitions are sparse. This cross-sectional study evaluates the outcome after resection of tarsal coalitions in children using the validated Oxford Foot and Ankle Questionnaire (OxAFQ).

Methods

Tarsal coalition patients between 5-16 years of age from Aarhus University Hospital (Denmark) and The Royal London Hospital (United Kingdom) were included. The patients were identified using patient and theatre register. All patients and proxies filled in the PROM: OxAFQ-C and OxAFQ-proxy respectively. The scores were calculated within each domain and reported as means (95% confidence intervals). Talocalcaneal coalitions were compared to calcaneonavicular coalition with regard to OxAFQ score and re-operation rate.

Results

27 patients and their proxies returned 54 questionnaires in total regarding 36 feet. Mean time from surgery to filling of the questionnaire was 25 (21-30) months. The relative mean OxAFQ score was higher in the School and Play and Emotional domain than the Physical domain, p = 0.007. The OxAFQ scores and re-operation rates were similar for both coalitions, p=0.63.

Conclusion

The OxAFQ PROM showed more encouraging results in playing or emotional health status than the physical health status. The outcome for both types of coalitions is similar.

Level of Evidence: IV

Introduction

Tarsal coalitions can cause pain and stiffness in the foot due to the bridge formation between the calcaneus and the navicular bone or talar bone.¹ Several surgical options have been explored. Although studies have indicated a positive outcome using arthrodesis,² recent studies show improved outcome with resection of the coalition and interposition of a biologic material, such as fat or muscle tissue, to prevent bridge relapse.^3–10 Consequently, arthrodesis should only be performed as salvage procedure, i.e., after failed resection of a tarsal coalition or in case of osteoarthritis in the transverse tarsal joint.^4,11,12

Outcome after coalition-resection have been widely described. However, the literature seems to lack validated and standardized reports on patient-reported outcome measures (PROM), leaving a gap for investigation.^9,10,13–15 Hence, much uncertainty remains regarding the long-term effect of treatment with resection of tarsal coalitions. The aim of this study was to report PROM data of the validated Oxford Foot and Ankle Questionnaire (OxAFQ) from the children and their proxies after resection of tarsal coalition and correlate the data gathered from the children to the data from their proxies.

Methods

The study was carried out in a cross-sectional retrospective design. Patients in this study were included at Aarhus University Hospital, Denmark AUH and The Royal London Hospital, United Kingdom RLH (Figure 1).

Figure 1.

Flowchart over inclusion. Inclusion and exclusion criteria as listed.

The Danish study population was identified by the Danish medical database using the International Classification of Disease 10 (ICD-10) diagnosis code for tarsal coalition (DQ668A). Patients with confirmed diagnosis and treated at AUH from 2006 until 2014 were included. Additionally, codes were used to identify patients with a tarsal coalition registered under a different diagnosis.

The British study population was identified using the orthopedic theatre lists from 2011-2014 at RLH.

First physician contact was defined as the first contact to an orthopedic specialist with the patient complaining from a symptomatic foot.

Data regarding type of coalition, operation date, applied imaging modalities, post-operative complications, additional surgery, co-existing disorders, age and sex were obtained by reviewing the patient charts.

Inclusion criteria were at least one symptomatic TC or CN coalition and surgical intervention with resection of the coalition performed between age 5 and 16 years at AUH or RLH. Exclusion criteria were presence of other foot and ankle disorders and learning disability that may impair the ability to respond to the PROM. All British participants received the child specific OxAFQ-C.¹⁶ The Danish population received a translated and validated version of the PROM.¹⁷ Both the child and their primary caregiver or proxy were asked to respond to the questionnaire. Contact with the patients was made through the phone. Patients with bilateral coalitions were asked to report data regarding the most symptomatic operated foot.

There are four domains in the OxAFQ: physical domain with a possible score of 0 to 24; school and play domain with a possible score of 0 to 16; emotional domain with a possible score of 0 to 16; final-item domain with possible score of 0 to 4. Higher scores are equivalent to better functional outcomes.

The outcome measures were the respective OxAFQ scores within each domain in the questionnaire, the number of additional surgical procedures performed, such as resection of relapsed bridge (RRB) or triple arthrodesis (TA) and the time from first physician contact to primary surgery.

Ethical consideration

The study was performed in accordance with the Declaration of Helsinki and was reviewed and approved by an internal review board and the regional ethical committee.

Statistics

Re-operation rate was calculated based on a total of 36 feet. The OxAFQ scores were calculated for each patient, based on the most symptomatic foot.

Both continuous and binary data were analyzed. Continuous data were time from surgery to PROM (OxAFQ) (months), age (years) and the individual OxAFQ score for children and their proxy (points). Binary data included were coalition type (CN or TC), the need for additional surgery due to unsatisfactory results from primary resection (yes or no), gender and observational period after initial physician contact to primary surgery (less than or more than 6 months). Values are reported as mean (95% confidence intervals). Fischer’s exact test was performed to test for significance in binary data. Pearson correlation analysis was performed for QxAFQ-C and OxAFQ-proxy.

All analyses were performed using STATA 13 (StataCorp. 2013 Stata Statistical Software: Release 13. College Station, TX: StataCorp Lp.).

Results

Thirty-six feet in 27 patients having a tarsal coalition (17 CN, 10 TC) were identified and included in the study (12 males, 15 females). All identified patients completed the OxAFQ-C and proxy. Patient characteristics and applied image modalities to diagnose the coalitions are given in Table 1.

Table 1.

Patient Characteristics and Imaging Modalities for Talocalcaneal and Calcaneonavicular Coalitions

	TC	CN	In Total
	(n = 10, feet = 13)	(n = 17, feet = 23)	(n = 27, feet = 36)
Patient characteristics
Males	4 (40%)	8 (47%)	12 (44%)
Females	6 (60%)	9 (53%)	15 (56%)
Mean age at surgery [years]	11.8 (10.5 – 13.0)	12.0 (10.5 – 13.7)	11.9 (11.0 – 12.8)
Mean time from 1st visit to surgery [months]	6.3 (4.3 – 8.3)	7.4 (4.2 – 10.6)	7.0 (4.9 – 9.1)
Mean time from surgery to OxFAQ [months]	25.1 (15.6 – 34.6)	25.2 (19.6 – 30.9)	25.2 (20.6 – 29.8)
Imaging modality
X-ray	1 (10%)	10 (59%)	11 (41%)
CT	5 (50%)	4 (23%)	9 (33%)
MRI	4 (40%)	3 (18%)	7 (26%)

Values are reported as mean (95% confidence intervals) or n (%).

The relative child and proxy OxAFQ-scores within the domains were highest in the School and Play and Emotional Domain, p < 0.1 (Table 2).

Table 2.

Mean Child and Proxy OxAFQ-Scores (95% CI) and the Number of Additional Surgeries Needed in Talocalcaneal and Calcaneonavicular Coalitions

	TC	CN	In Total
	(n = 10)	(n = 17)	(n = 27)
OxAFQ-C
Physical domain	15.5 (11.1-20.0)	17.6 (14.3-21.1)	16.9 (14.3-19.4)
School & play domain	11.3 (8.0-14.56)	13.8 (11.9-15.8)	12.9 (11.2-14.6)
Emotional domain	12.6 (10.4-14.8)	14.6 (12.9-16.3)	14 (12.6-15.2)
Final item domain	2.3 (1.6-3.0)	2.6 (1.8-3.4)	2.5 (2.0-3.0)
All domains	41.7 (32.8-50.6)	48.6 (41.4-55.9)	46.1 (40.6-51.5)
OxAFQ-proxy
Physical domain	13.5 (8.11-18.9)	16.6 (13.4-19.9)	15.5 (11.5-19.5)
School & play domain	11.5 (8.0-15.0)	12.6 (10.5-14.8)	12.2 (9.6-14.9)
Emotional domain	12.6 (9.4-15.8)	13.3 (11.3-15.3)	13.0 (10.6-15.5)
Final item domain	2.1 (0.9-3.3)	2.3 (1.6-3.0)	2.2 (1.3-3.1)
All domains	39.7 (27.6-51.9)	44.9 (37.4-52.4)	43.0 (36.9-49.0)
Surgery
Additional surgery	2 (20 %)	3 (18 %)	5 (19 %)

Values are reported as mean (95% confidence interval).

The collected child-questionnaires gave a mean OxAFQ-C-score of approximately 70% of the possible maximum score for all four domains for the TC coalitions. For the CN coalitions, the mean OxAFQ-C-score was 81% for all domains. The combined OxAFQ-C and Proxy score was similar for both TC and CN coalitions (Table 2). OxAFQ scores did not increase with time from surgery to completing the questionnaire for both TC and CN coalitions, r = -0.27; -0.02 (Figure 2).

Figure 2.

Scatterplots for the combined OxAFQ-C and Proxy score for Talocalcaneal (TC) and Calcaneonavicular (CN) coaltions. Time from surgery to completing the questionnaire is represented on the x-axis and the combined OxAFQ score for all domains is showed on the y-axis.

A: Pearson’s coefficient = -0.14 (CN), Pearson’s coefficient = -0.07 (TC)

B: Pearson’s coefficient = -0.02 (TC), Pearson’s coefficient = -0.27 (TC)

The collected proxy-questionnaires demonstrated a similar mean OxAFQ-proxy-score within each domain compared to the OxAFQ-C with a strong correlation, r = 0.68 – 0.89 (Figure 3). The mean TC OxAFQ-proxy-score was 66% for all domains, and 75% for the CN coalitions. The re-operation rate was 19% and similar for both types of coalitions, p=0.63. No significant difference was observed in the OxAFQ scores between the re-operated and primary resected coalitions.

Figure 3.

Scatterplots for the Physical, School and Play and Emotional Domain. The OxAFQ score for proxies appears on the x-axis with the matched OxAFQ score for children on the y-axis. Linear regressions have been performed accordingly.

A: Physical domain. Pearson coefficient = 0.89

B: School and Play domain. Person’s coefficient = 0.69

C: Emotional domain. Pearson’s coefficient = 0.68

The mean time from surgery to filling in the questionnaire was also similar for both coalitions (Table 1).

Discussion

Most studies on outcome after resection of tarsal coalitions have reported clinical outcomes rather than patient reported outcomes.^{5–8,13,18,19} Although a few recent studies have dealt with this issue using PROMS,^9,10,14 no studies have utilized a validated child-specific questionnaire. In this study we have quantified the obstacles and their influence on daily life using the validated child-specific OxAFQ-C. The OxAFQ scores were similar for both coalitions and highest within the School and Play and Emotional domain. Previous studies have advocated a more favorable outcome for CN coalitions,^18,20 however the PROM scores were similar between the two coalitions in our study, which also reflects with recent other reports.^{4,10,14,21,22} The higher relative scores within the School and Play and Emotional domains (maximum of 16 points) indicate that the children are not impaired in the daily play or affected emotionally despite the lack of satisfaction of the physical outcome. Data collected from the children seems to be in accordance with the parental observations as shown in Figure 3. The strong correlation between the data collected from the children and proxies suggests it is reliable to gather the information from the proxy. A possible bias in the collected data may be the time from surgery to completing the questionnaire. However, as seen in Figure 2, there does not seem to be any correlation between the type of coalition, the time from surgery to questionnaire and the OxAFQ for both proxies and children. In fact, the Pearson’s correlation coefficient between -0.27 and -0.02 suggests a negligible correlation.

Despite the overall positive outcome upon resection in recent studies, some patients continue to have a symptomatic foot. This raises the issue of need for additional management as the outcome for conservative treatment options differs somewhat with studies reporting a success rate of 25-30% on the short term whilst the long term outcome after 5 years was 74% using conservative treatment.^23,24 An estimation of 25% of all coalitions are symptomatic, leaving the vast majority to be a random non-symptomatic find.^25,26 A prospective study with OxAFQ scores pre- and post-operatively can be performed to clarify this important issue and the improvement or worsening of the condition of the patients after resection of tarsal coalition.

Limitation to this study is the retrospective cross-sectional design. First, no pre-operative data exists. Despite the focus of this study being to report data from a validated PROM postoperatively, it could have been advantageous to correlate these data with pre-operative PROM data to investigate improvement or worsening of the condition. Moreover, several bilaterally resected coalitions are present in this study, whilst only the most affected foot was reported in the OxAFQ-C and proxy. This may suggest the presence of a slightly higher OxAFQ-core if a questionnaire was completed for each foot, especially in the physical domain, which scored lower than the two other main domains in relative score. Another limitation is the study population. To achieve a large study population, patients from AUH in Denmark and RLH in the United Kingdom were enrolled. The registration procedure of patients at RLH differed from AUH. Therefore, no search could be done using the ICD-10 code. Instead, a manual search of the theatre lists was performed. This approach of manually identifying patients retrospectively, is more likely to miss some cases than the Danish counterpart, where we were able to search for diagnosis and operation codes. In an attempt to locate some of the possibly missing patients, personal operation lists provided by the consultants were also applied.

In conclusion, children and proxies can receive valid and specific information about the patient reported outcome after surgical resection, with no difference being detected between TC and CN coalitions. Despite the persistent physical affection, the children are less affected emotionally and in their daily play. Notably, OxAFQ-C and proxy were coherent and independent from the time from surgery to completing the questionnaire.