The procedure is performed under general anesthesia with the patient in the supine position. A hand table is typically used for a larger child, while an arm board or just the operating room table may be used for a smaller child. A small arm tourniquet is applied and the extremity is prepared and draped as per routine. Intravenous antibiotics should be administered within 60 min of the skin incision. Bruner incisions are preferred over midaxial incisions, especially in infants and children less than 5 years of age. Kavouksorian and Noone (1982) reported that in younger patients, midaxial incisions may migrate palmarward and result in digital flexion contractures. Both digital neurovascular bundles should be surgically explored. In contrast to the adult patient, it can be very difficult to assure the integrity of neurovascular structures based solely on the preoperative physical examination of the pediatric patient.

The tendons are subsequently explored. Windows can be made in the pulley system as necessary. The site of pulley system “windowing” should be made after careful assessment of the probable repair location. The surgeon should not open the pulley system in the same location for every case. For instance, if the tendon laceration is very distal, the A4 pulley may need to be sacrificed, but the A3 pulley may be preserved. Venting of the entire A4 pulley, or part of the A2 pulley (up to 50 %), may be required and should still allow for an acceptable result (Kwai Ben and Elliot 1998; Tang 2013).

Surgical repair techniques vary depending on the author and center. The authors’ preferred suture configuration repair method for both adults and those older children with larger tendon sizes is the cross-locked cruciate–interlocking horizontal mattress (CLC-IHM) method. Both the core component of the cross-locked cruciate and the circumferential component of the interlocking horizontal mattress have superior biomechanical characteristics when compared to other commonly used methods (Figs. 1 and 2) (Croog et al. 2007; Lee et al. 2010; Lee 2012; Vigler et al. 2008). Unfortunately, there is not typically enough space in the tendon for this repair technique in small children under the age of 10 years.

In children under 2 years of age, the FDP is reported to measure 2–3 mm in width and 0.5–1 mm in thickness. For this age group, most authors use a two-strand modified Kessler suture configuration (Fig. 3) (Berndtsson and Ejeskar 1995; Elhassan et al. 2006; Fitoussi et al. 1999; Grobbelaar and Hudson 1994; Herndon 1976; Kato et al. 2002; O’Connell et al. 1994). In children over 5 years of age, a four-strand repair (Fig. 4) may be feasible depending on the size of the child (Elhassan et al. 2006; Navali and Rouhani 2008; Nietosvaara et al. 2007). The rupture rates for four-strand repairs have been shown to be lower than those observed following two-strand repairs (Navali and Rouhani 2008; Nietosvaara et al. 2007).

Al-Qattan advocates using three “figure of 8” sutures for children regardless of age. He reported higher tensile strength than the modified Kessler configuration with this technique (Al-Qattan and Al-Turaiki 2009). The chief disadvantage of the “figure of 8” repair is that the resulting repair site is bulky and oftentimes too large to freely pass underneath the flexor sheath. Therefore, venting of the flexor pulley system is almost always required. Most authors supplement the core suture with a running circumferential suture, which is still surgically practical in the pediatric patient despite the small size of the tendons. Isolated repair of the FDP, as compared to repair of both the FDP and FDS tendons, is controversial in both the adult and pediatric literature. Some authors (Al-Qattan 2013; Fitoussi et al. 1999; Kato et al. 2002) recommend only repairing the FDP tendon injury, whereas other authors (Grobbelaar and Hudson 1994) advocate repair of both. Paillard and colleagues (Paillard et al. 2002) showed that resection of one slip of the FDS tendon decreased gliding resistance as compared to pulley plasty in a cadaveric model. The subject of repair of both slips of FDS tendon, versus one slip, as compared to none is also controversial. Based on the above study and clinical experience, it is reasonable to repair one slip of FDS tendon provided that the FDP tendon still glides easily after repair. It is imperative that the tendons glide with no interruption in the operating room. It is a fallacy to think that the gliding will improve with time and motion. It is safe to conclude that the motion the surgeon achieves in the operating room is the best that will be achieved postoperatively.

In regard to the core suture purchase length, it has been shown in adults that the optimal suture purchase length biomechanically is between 7 and 10 mm (Lee et al. 2010; Tang et al. 2005). For children, Navali and Rouhani (2008) recommend the rule of a suture purchase length of 1.5–2.0 times the width of the tendon. In a series of 12 children with ages between 15 and 23 months, Al-Qattan (2013) showed that the average width of the FDP tendon in Zone II was 2.5 mm. Depending on patient size, the suture purchase length should typically equal 3.5–4.5 mm in this patient population. Navali and Rouhani (2008) also advocated that this distance should not exceed 5 mm.

Elhassan et al. (2006) reported on the Bunnell pullout repair (Bunnell 1948) tied over a button on the nail plate (Fig. 5) in a series of 16 children. One repair ruptured postoperatively, while the remaining cases had a TAM (total active motion) of 89 % of normal which is rated as excellent (range 66–100 %). Al-Qattan (2013) advocates avoiding the possible complication of infection and nail deformity associated with the dorsal button technique by repairing the proximal tendon to the short distal stump and reinforcing it with a repair to the volar plate of the DIP at a point distal to the joint. Special care is required to avoid placing sutures deep into the physis, which may result in a growth disturbance. Al-Qattan (2012) reported on the use of this technique in a series of 10 children. There were no cases of ruptures or growth plate injury, and all patients had excellent results on the Strickland–Glogovac (1980) score. Using the Moiemen–Elliot (2000) scale, excellent results were achieved in 5 children, good results in 1, and fair results in 4.

Repair principles and outcomes are similar in the flexor pollicis longus (FPL) as compared to the FDP tendons. Grobbelaar and Hudson (1995) reported on nine children with FPL repairs managed postoperatively with controlled mobilization. Good to excellent results using Buck-Gramcko scores were achieved in seven patients, with no cases of tendon rupture reported. Fitoussi et al. (2000) reported on 16 children with FPL repairs using a modified Kessler suture repair technique. Buck-Gramcko evaluation showed good to excellent results in all but one patient who had a tendon rupture post-repair. One third of the cases had at least a 30° decrease in ROM as compared to the normal side. The authors also showed that digital nerve injury had no negative effect on eventual outcome. Overall, the authors had improved results with earlier mobilization. Orhun et al. (1999) had excellent Buck-Gramcko evaluations in 6 of 7 patients, with one remaining good result in a series of pediatric Zone II FPL repairs. In the setting of a chronic FPL injury, good results have been reported with a two-stage tendon transfer utilizing a silicone tendon implant followed by a transfer of the FDS of the ring finger (Yamamoto and Fujita 2013).