Abstract
The outcome of total wrist arthrodesis was reviewed in 36 patients with osteoarthritis after a minimum follow-up of 4 years. Pain relief was not complete, and although 20 were pain free at rest, only six were pain-free during manual activity. Grip strength was 63% of the contralateral side and the DASH score remained high. Only 21 of the 34 could be re-employed. The mean time off work was 14 months. Complications were numerous and additional surgery was required in 21 patients.
INTRODUCTION
For many years total wrist arthrodesis or radiocarpo-metacarpal arthrodesis was considered the gold standard for unresolved wrist problems. For most insurance companies this situation was favourable since total wrist fusion was then the “administrative” end of a sometimes long history. There are however more critical voices (Gaisne et al., 1991; Nagy and Buchler, 1998). The purpose of this survey is to evaluate critically the outcome of total wrist arthrodesis in a community with a well developed and protective social security system.
PATIENTS AND METHODS
Population
We reviewed 36 patients who had undergone total wrist arthrodesis for osteoarthritis of the wrist (24 men, 12 women: mean age 42 years, range 24–65 years). The mean follow-up was 7 (range, 4 to 9) years. The cause of the osteoarthritis is detailed in Table 1.
Twenty-four patients underwent surgery on the right wrist of which 23 were dominant. Previous operations had been performed in 20 patients: ten had undergone one operation, eight had undergone two operations and two had undergone three operations. The exact nature of these operations was not always clear. Most of these patients were referred by other physicians.
Surgical technique
Two different techniques were used. The wrist is approached dorsally between the third and fourth extensor compartments. The articular cartilage of the radiocarpal, midcarpal and carpometacarpal joint is removed. In the first group (24 cases) a 2 cm wide slot was created from the radius to the base of the third metacarpal and a corticocancellous iliac graft was harvested and placed in this and fixed with screws (Fig 1) (Rayan, 1986). In the second group (12 cases) a prebent AO/ASIF plate was used and cancellous grafts were interposed in the denuded joints (Fig 2). Since this plate became available, the former technique was abandoned.
In 12 cases a Darrach resection of the distal ulna was performed simultaneously. Postoperatively a plaster cast was worn for 6 weeks. Finger mobility exercises were started immediately.
Evaluation
The patients were evaluated by an independent observer (J.T.). Twenty-eight were seen for questioning and a physical examination and eight were interviewed by phone only. All patients completed a DASH score (Hudak et al., 1996) and estimated their residual pain on a Visual Analogue Score (VAS). Several correlations were computed using Pearson’s correlation coefficient.
RESULTS
Pain at rest resolved completely in 20 patients but only six were fully painfree during activity. The mean VAS for pain at rest was 2.5 (range, 0–8) and it was 5.4 during activity (0=no pain, 10=severe pain). The pain usually was ill-defined. In seven cases the pain was obviously due to instability of the proximal ulnar stump. When we analysed the cases with severe residual pain, we found that the fixation plate was prominent in one case, and one case each of extensor tendinitis, reflex sympathetic dystrophy and nonunion of the middle finger carpometacarpal joint. No explanation for the residual pain was found in any of the other cases. Comparison of the painful and pain-free cases revealed no differences in gender, age distribution or work. Patients with a Darrach procedure had no more pain than those without (VAS painscores during activity=5.52 (SD, 2.91) and 5.24 (SD, 3.02), respectively).
The mean DASH-score was 44 (SD, 22); (0=disability, 100=fully disabled).
Patients who had suffered a work-related accident had a DASH score of 44 (SD, 23) compared with 45 (SD, 22) for the others (P=0.67). When the total wrist arthrodesis was performed as the first operation the DASH score was 51 (SD, 21; N=16) compared with 40 (SD, 20) for those who had undergone previous operations (P=0.14).
The mean period of incapacity in those patients who returned to work was 14 (SD, 11; range, 4–48) months. Only 11 patients returned to their previous job (mean DASH, 33; SD, 18), ten had to change to lighter work (mean DASH, 38; SD, 22) and 14 could not be re-employed (mean DASH, 56; SD, 20) (Table 3).
Grip strength was 63% (SD, 22) of the contralateral side (N=28). The pre-operative grip strength (N=16 patients) was 53% of the contralateral side, but this difference is not significant. The patient’s subjective evaluation was excellent for 5, good for 15, fair for 13 and poor for 3. The grip strength was 65% (SD, 21) of the contralateral side when a Darrach’s procedure was performed, compared to 60% (SD, 21.79) for the others (P=0.53).
Complications were numerous. In the first 3 months postoperatively we observed one plate loosening, three infections of the bone graft donor site, one infection of the wrist and one reflex sympathic dystrophy. Late complications included ulnar wrist pain (seven cases: five with the ulnar head preserved and two after a Darrach’s resection), radiocarpal nonunion (two cases), extensor tendinitis (two cases), failure of the metal plate (one case) and intrinsic tightness (one case). Several secondary operative procedures were necessary in 21 patients (Table 2): 12 had one procedure, eight had two procedures and one had three procedures. Removal of the metal plate was usually performed at the request of the insurance company and not because of tendon irritation.
There were two symptomatic pseudarthroses of the middle finger carpometacarpal joint, both in cases treated with an iliac crest bone block.
There were no significant differences between the two techniques although the statistical power of this study is weak. One radiocarpal pseudarthroses was seen in each group strength.
DISCUSSION
The outcome of wrist arthrodesis has been studied by several authors. Their results and conclusions vary widely and the outcome probably depends on the availability of socio-economic support, the composition of the patient cohort and the outcome assessment used in the study.
Complete pain relief after total wrist arthrodesis has been reported in 100% (Weiss and Hastings, 1995), 84% (O’Bierne et al., 1995), 76% (Houshian and Schrøder, 2001) and 70% (Sagerman and Palmer, 1996) of patients. However other series reported less optimistic results: Nagy and Buchler (1998) found that only 56% of their patients had pain relief and only 25% of Gaisne et al.’s (1991) patients were painfree. Field et al. (1996) found a postoperative pain VAS of 4 and Sauerbier et al. (2000) reported that pain was reduced on average by half. The postoperative DASH scores were 46 and 52 in two recent series, indicating some persistent problems (Kalb et al., 1999; Sauerbier et al., 2000). Of interest is that our outcome of wrist fusion was not influenced by whether or not the accident occurred at work.
The mean duration of time off work after total wrist fusion has been reported at between 4 and 6 months (Field et al., 1996; Sauerbier et al., 2000; Weiss and Hastings, 1995) but 15% to 54% of these author’s patients did not return to work. Gaisne et al. (1991) reported a mean time off work of 15 months and found that only seven of his 34 patients could resume their previous work. Seven obtained lighter work and 17 became unemployed (Table 3).
The impairment rate after total wrist fusion is reported at 5% to 36% (Hastings et al., 1996; Houshian and Schrøder, 2001, Sagerman and Palmer, 1996; Weiss and Hastings, 1995). Our findings are very similar, but social security care in Belgium is good which probably explains the long period of incapacity. The impairment in our country attributed to a wrist fusion without complication is 15%.
Our results suggest that total wrist arthrodesis is not the ultimate solution to complex wrist problems. Many patients need supplementary operations, the DASH score remains moderately high and only half of our patients were satisfied with their outcome. Return to work is not universal and the period of sick leave is long.
However, in all of these cases no alternative options to total wrist fusion were judged reasonable and in our opinion a total wrist arthrodesis was the only possible solution.