Abstract
Introduction:
Cleft lip and palate deformities are one of the most common birth defects. The alveolar cleft requires bony repair to allow proper eruption of dentition. The purpose of this study is to evaluate success in the repair of alveolar clefts with iliac bone grafts.
Aim:
The aim of this study is to restore the function and form of both arches with a proper occlusal relationship and eruption of tooth in the cleft area.
Subjects and Methods:
Five patients were selected irrespective of sex and socioeconomic status and whose age was within the mixed dentition period. The iliac crest is grafted in the cleft area and subsequently evaluated for graft success using study models, periapical, and occlusal radiographs.
Results:
At the time of evaluation, teeth were erupted in the area and good alveolar bone levels were present. Premaxilla becomes immobile with a good arch form and arch continuity. There are no major complications regarding pain, infection, paresthesia, and hematoma formation at donor site without difficulty in walking. There is no complication regarding pain, infection, exposure of graft, rejection of graft, and wound dehiscence at the recipient site except in one case.
Conclusions:
Long-term follow-up is required to achieve maximum advantage of secondary alveolar grafting; the age of the patient should be within the mixed dentition period, irrespective of sex and socioeconomic status. It may be unilateral or bilateral.
Keywords: Bone grafts, iliac bone grafts, secondary alveolar clefts
Introduction
Secondary alveolar bone grafting (ABG) is performed to allow proper eruption of the canine teeth through the cleft segment and to close the oronasal fistula.[1,2] Successful ABG is important to allow for orthodontic positioning and retention of the cleft adjacent teeth as occlusion is optimized. Most commonly, the source of bone is the iliac crest. There ported success rate of unilateral cleft repairs is up to 90% and <80% in bilateral alveolar cleft repairs.[3] Conflicting claims have been made with regard to success and time of surgery in this area. Prominent among these have been the studies of Bohn, Bjork and Skieller, Waite and Kersten, and Boyne and Sands. Fresh autogenous bone is the ideal bone graft material because it supplies living immunocompatible bone cells essential to osteogenesis.[1,4,5,6,7] The present thinking that alveolar grafting should be ideally somewhere between 6 years of age and before eruption of teeth in cleft margin.[8]
Subjects and Methods
The present study includes the management of five patients having cleft alveolus (including of 2 years follow-up). All patients were unilateral clefts. Patients’ selection was done irrespective of sex and socioeconomic status, and all patients were above 5 years, previously operated cases were not included. Extraoral and intraoral examinations were carried out for all the patients.
The cases were diagnosed on the basis of clinical and radiological (intraoral periapical and intraoral occlusal views) features; unilateral cleft lip and palate patients were included in the study [Figures 1 and 7]. Oral prophylaxis was carried out in every case to improve the oral hygiene. Restoration of carious teeth was done in cases wherever required.
Figure 1.
Preoperative clinical picture of the alveolar clef
Figure 7.
Pre operative occlusal Radiograph
Operative procedure
The patient was suitably positioned for this procedure under nasoendotracheal intubation in general anesthesia. The surgical site was prepared using the standard methods. Incisions were made along the cleft margins splitting the labial/nasal and palatal/nasal mucosa. Labial incision was extended for adequate exposure. Two gingival flaps were then raised. Palatal mucoperiosteal flaps were elevated just enough to visualize the cleft [Figure 2]. All hypertrophic mucosae (if present) at the cleft edge were excised to allow accurate approximation. The flaps were sutured together to form a nasal floor at the same level as on the noncleft side with interrupted sutures (4-0 catgut).
Figure 2.
(a) Visualization of the cleft. (b) Buccal and palatal flap reflection
The palatal mucoperiosteal flaps were then approximated medially with interrupted 4-0 absorbable sutures, thus creating a soft-tissue pocket to accommodate the bone graft. The corticocancellous bone graft harvested from the iliac crest bone was packed tightly to completely fill the bony cleft and to restore the thickness and height of the nasal floor and the maxilla as close to normal as possible. Any permanent teeth erupting through the cleft was covered with bone graft [Figure 3].
Figure 3.
Corticocancellous iliac bone graft placement
Finally, the previously raised gingival mucoperiosteal flaps were sutured together and inferiorly to the palatal flaps to provide complete coverage of the bone graft [Figure 4]. The flaps were further secured with a few interrupted sutures between the flap and papilla, while the area of the back cut posterior was left open to heal secondarily. Corticocancellous graft from the iliac crest was harvested by standard surgical methods.[9]
Figure 4.
Closure done with 3-0 vicrly
Harvesting the iliac crest graft
The iliac crest was used as a donor site for harvesting the graft. The patient was placed in the supine position. A sandbag was placed beneath the hip to elevate and slightly rotate the anterior iliac crest. The surgical site was prepared by standard methods. The landmarks for the anterior iliac crest surgery include the anteroposterior iliac spine and the iliac crest as it curves superiorly and posterior from the spine.
An incision was made through the skin and periosteum, starting approximately 1.0 cm lateral and inferior to the anterior iliac spine for 6–8 cm. Before giving the incision, the skin was retracted so that the incision lied lateral to and below the crest instead of over the crest. The superior portion of the incision was then retracted medially to expose the crest of the ilium. The abdominal musculature was separated from the iliac crest, and the inner surface of ilium was freed from soft tissue using a periosteal elevator [Figure 5]. A block of corticocancellous bone was taken from the lateral surface of the iliac crest by first making two parallel 2–4 cm vertical cuts with a chisel. The inferior portion of these vertical cuts was connected by a horizontal cut with oscillating saw or osteotome. A straight orthopedics chisel was used to separate the graft from the lateral cortex. Periosteum of the crest was sutured tightly to the abdominal muscles with 2-0 chromic catgut. Suction drain was fixed, and the remaining layers fascia and fat were closed with 2-0 chromic catgut. The skin was then closed with a 3-0 prolene/nylon subcuticularly [Figure 6]. Pressure dressing was kept over the wound to prevent superficial hematoma formation. Appropriate postoperative antibiotics were given for 5 days. Drain was removed on the second postoperative day, and sutures were removed on the 10th day of the surgery. To reduce the period of discomfort and disability, early ambulation was advised. The patients were advised to apply framycetin ointment (Sofradex) over the suture line. Postoperative complications were evaluated weekly at the donor site as pain, infection, paresthesia, hematoma formation, and difficulty in walking (change in gluteal gait), whereas at the recipient site as pain, infection, exposure of graft, rejection of graft, and wound dehiscence.
Figure 5.
Incision given on the anterior iliac spine
Figure 6.
Closure done with 3-0 prolene
Results
In our study, patients were followed up for 18 months for two patients, 12 months for two patients, and 6 months for one patient. According to Bergland's criteria, two patients had more than three-quarters of the normal bone height when assessed. Other two patients had less than three-quarters of normal bone height, and one patient had failure [Table 1]. The assessment was carried out in occlusal radiographs [Figures 7 and 8]. In the recipient site, there were pain, slight ooze, and mild swelling postoperatively for a period of about 24 h in two patients, whereas there were no complaints postoperatively in the other two patients. In one patient, there was complete graft failure in 1½ years period [Graph 1]. Postoperatively, two patients had pain at the donor site which was relieved with use of analgesics. Three patients did not have any donor site complications. Postoperatively, ambulation occurred at an average of 4 h (4–6 h). All patients reported cutaneous anesthesia in the region of surgical site and experienced only mild discomfort on ambulation. Two patients were discharged the day after surgery, and two patients were discharged 2 days after surgery. During the postoperative course, no complications such as hematoma, seroma, paresthesia, or infection were experienced in the donor site. Although the patients favored the operated site during the ambulation for the first postoperative week, none showed a gait disturbance in the follow-up visits. Blood loss from the iliac crest harvest averaged 110 mL (range: 80–120 mL). The average length of the cutaneous incision was 27 mm (range: 21–40 mm). The harvested bone was condensed in a wet gauge to increase the density. The volume measured was 4 mL (range: 3–5 mL). The volume of bone harvested was more than sufficient in all cases [Table 2].
Table 1.
Bergland's criteria
Figure 8.
Post operative and postoperative radiographs
Graph 1.
Pie Diagram showing Success rate of the graft in the cleft defect
Table 2.
Size of the Cleft defect and volume of the graft placed
Discussion
The transplantation of a piece of bone from one part of the body to another to repair a skeletal defect. Several types of bone grafts are available for use in reconstructive surgery. A successful classification categorizes the bone grafts according to their origin and thus their potential to induce an immunological response. ABG is frequently performed in the treatment of patients with alveolar and palatal clefts. The principal purpose of this procedure is improvement of the dental alignment. Repair of the alveolar cleft with bone grafting enables eruption of the canine through the grafted bone or placement of the implants. Alveolar repair should not impede maxillary or midfacial development.[10] The early repair (before 2 years) would improve results in nutritional, esthetic, and psychological development. On the contrary, studies showed that the results were not as good as expected. Early repair lead to midfacial growth problems from extensive hard palatal dissection around and across the vomerine bone-premaxillary suture.[11,12] The indications of secondary alveolar cleft grafting are grafting during mixed dentition (after eruption of maxillary central incisor and before the eruption of the canine, the root of which is one-third formed).[10] The advantages of secondary ABG are early stabilization of maxillary arch, obliteration of oronasal fistulas, provide bony support for the eruption of the canines and the orthodontic alignment of the teeth.[11,13,14,15] Secondary alveolar bone grafting would obtain a complete osseous arch and superior results in facial contour, as it was advocated that primary grafting was disturbing the sagittal and vertical growth of the maxilla, maximum results were obtained by grafting between 9 and 11 years, just before full eruption of the canines (Boyne and Sands).[1] Most surgeons prefer to perform grafting when the adjacent unerupted canine root is one-fourth to two-third complete. Actual studies show that the best results are obtained during the premature canine dentition period. Today, secondary ABG is the treatment of choice.[10] ABG varies according to the timing of the surgery (1) primary grafting: Before 2 years, (2) early secondary grafting-between 2 and 5 years, (3) secondary grafting-between 5 and 16 years, and (4) late secondary grafting: above 16 years. The best results will be achieved when the graft is inserted at the age of 9–11 years before the eruption of the canine.[11,12,17,18] Complications of secondary alveolar cleft grafting are partial or complete loss of bone graft, wound dehiscence, infection, residual oronasal fistula, lack of adequate periodontal bone or soft tissue support, soft-tissue necrosis, external root resorption, failure of tooth in area to erupt spontaneously, donor site morbidity, collapse of dento-osseous segment, loss of vestibular depth, inadequate attachment of gingiva to teeth in bone graft area, prolonged period of disability, adverse effect on patient and family, psychological well-being, etc.[12,19] In a study done by Sadove and Eppley on comparing iliac and calvarial bone grafts, the iliac bone graft success rate was high at 93% with 14 out of 15 grafts; however, in our present study, the success rate was 80% with 4 out of 5 grafts.
Secondary grafting can be done with three kinds of materials: autogenic, allogenic, and alloplastic. It is generally agreed that the particulate materials are preferred to block corticocancellous materials because they can be incorporated faster and teeth erupt into the graft. Iliac bone is the main donor site choice today; however, because of morbidity (hematoma, infection, more blood loss, deformity, longer operation, and hospitalization) new materials are being studied. Complication rates of over 15% have been reported by Swan and Goodacre. In the 1950s, frozen allogenic bone in alveolar cleft management was reported to have a superior result (Black Dhal et al.). Particulate allogenic bone grafts was used in oronasal fistulas for the closure of the fistula and movement of unerupted teeth to the graft. Clinical and radiological investigations indicated bony bridging; however, a difference in the density ensured that it should be used in unilateral cleft patients without unerupted teeth. In these cases, orthodontic manipulations can be done 3 months postoperatively. It has advantages such as ease in obtaining and sterilizing and preserving its chemical and structural properties for a long. The most important advantage is that it eliminates the need for a second operation site, eliminates morbidity, and shortens the hospitalization time. The disadvantage of this material is increased time for incorporation, suspicion for orthodontic teeth eruption, and its periodontal support.[12] According to Turvey et al., allogenic particulate materials can be used as grafting material and that orthodontic treatment should be started in the postoperative 6th month when the normal trabecular pattern is obtained radiologically. Marx et al. compared autogenic and allogenic grafts in animal studies and both groups achieved bony bridging [Graph 1]. However, the radiographic evaluations showed that 30% of the cross-sectional area of the cleft was occupied by viable bone. Nique showed that the autogenic bone grafts stay radiolucent to 6 months postoperatively during the remodeling period. However, on the contrary, allogenic grafts are still radiopaque in that period because their resorption takes more time. Actually, the most suitable material for alveolar grafting is autogenic bone. Membranous bone grafts were revascularized earlier, and grafts survived more than endochondral bone grafts (Bilkay et al.).[12] Donor bone selection should be done according to the patient, age, and volume of the defect. Donor sites are ribs, tibia, calvarium, and mandibular symphysis. Calvarium has advantages such as an no visible scar, no secondary deformity, an abundance of bone in children, less postoperative pain, briefer hospitalization, and donor site in the same operative field and greater graft survival with membranous bone. Successful results were obtained by Wolfe; however, Jackson reports showed that the poor results because of dissatisfaction with the ossification of the graft and periodontal defect along the adjacent teeth.[12] The success rate of calvarial bone grafts done by Hudson brace technique showed a success rate of 80% and by calvarial craniotomy technique showed a success rate of 50%.[20] Mandibular symphysis has been used as a membranous bone donor site with minimal morbidity and satisfactory success rate of the grafted cleft. The main problem was the amount that can be harvested is not always sufficient as in cases of bilateral clefts. Split-rib graft used in primary bone grafting provides a continuous alveolar arch; however, tooth eruption and healthy periodontal support are doubtful and graft incorporation is also less sure. Pickerell showed that teeth do not erupt into the rib graft. The sites implanted with hydroxylapatite in animal models with surgically achieved clefts formed a generalized; however, variable fibroblastic response nears the interface between the wall of defect and the hydroxylapatite granules 3 months later. At 6 months, most hydroxylapatite granules were covered with maturing and well-collagenized fibrous tissue sheath. Partial bony bridging and some giant cells around the granules were noted. Three months after autogenic bone grafting, the new bone was largely woven in nature and consisted of immature fibrous connective tissue (Cullum and Horswell). Materials such as hydroxylapatite are used in adult patients to augment the alveolar contour, but not in patients having unerupted teeth.[12] Autografts or allograft materials are used for most often for these procedures. The current gold standard is the autologous bone graft commonly from iliac crest, tibia, and cranial bone. The anterior iliac crest provides abundant cancellous bone and is ideal for condensing into alveolar clefts, greater volumes of the grafts is obtained, two surgical teams can work at the same time and hence, the operation time is shorter and does not cause severe complications and minor problems can be reduced with careful manipulations.
Conventional open approach for harvesting iliac bone graft requires skin incision, muscle dissection, and periosteal elevation. These can result in prolonged pain and limping, infection, hematoma, irritating scar, sever pain from the pressure of cloths and belt, and local hypoesthesia or anesthesia with the injury of lateral femoral cutaneous nerve. To reduce these complications, bone marrow biopsy needles were used which provided superior results regarding the donor site infection, hematoma, sensory disturbances, and contour deformities. In a study done by Rawashdeh, among 64 patients, 59 (95.3%) patients suffered pain at the hip donor site with three patients (4.7%) reporting no pain. The present study had two patients complaining of pain (40%) and three patients reported no pain (60%). The same study also reported 58 (91%) patients were able to walk within 24 h, whereas in our study, all the patients were able to walk within 24 h postoperatively 4 (80%) patients were able to walk within 12 h postoperatively, and 1 (20%) patient walked 16 h postoperatively. The average duration of the limp was 5.4 days as stated in the same literature which was supportive to our results also, in which patients had a slight limp in the 1st postoperative week which resolved by the end of the 2 weeks, and patients had normal gait. Swan and Goodacre's study reported a single case of persistent numbness of the scar and another case of hyperesthetic scar and one case of single hypertrophic scar.[21] In our study, the scar length in the donor site was found to be around 3–5 cm in the followup period. None of the patients reported numbness, hyperesthesia, or hypertrophy postoperatively. Complex methods of bone graft analysis have been proposed; however, the Bergland grading system of measuring bone graft success has remained the gold standard and is most often that of the normal side and a four-point scale is used to categorize each graft. Here, not only the occlusal level of the bone grafts but their basal levels are also evaluated. The occlusal level corresponds to the level of the amelocemental junction of the cleft teeth. Native alveolar bone surrounds the root of the cleft central incisor and much of the cleft canine. In this, the basal level of the graft is also recorded and compared with that of the normal interdental bone. The latter corresponds to the level of the root apices and anterior nasal spine. Once the occlusal and basal levels of the grafts are evaluated total bone graft height at the mid-region of the cleft is measured.[14,22] In our study, patients were followed up for 18 months for two patients, 12 months for two patients, and 6 months for one patient. In a study conducted by Hynes and Earley, using Bergland criteria, cases were assessed were in, out of 71 grafts, 43 (60.5%) and 23 (32%) were characterized in type I and type II, 4 (6%) were in type III, and 1 patient had complete graft failure. 14 The present study had 2 (40%) patients have more than three quarters of the normal bone height when assessed. Other 2 (40%) patients had less than three-quarters of normal bone height and 1 (20%) patient had failure [Graph 1]. The assessment was carried out in occlusal radiographs. In all our cases, the size/volume of the cleft area measured from the occlusal radiograph and transposed on to the graph sheet. The number of the square centimeters occupied by the graft is calculated. This is useful for measuring volume of the iliac graft to be taken and compare with the postoperative radiograph to measure the amount of bone generated. In our study, the size of the defect ranges from minimum to maximum (0.8–1.8 cm2), the volume of the graft taken range from 3 ml to 4.5 ml. The postoperative radiograph showed 40% of the cases (2 cases) showed good amount of bone. Another 40% of cases (2 cases) showed optimal amount of bone formation and 20% (1 case) failure seen. Fresh autogenous iliac cancellous bone for secondary ABG provides live, immune-compatible bone cells essential for osteogenesis which promises good results. Our study is in accordance with other studies.
Conclusions
The advantages of secondary ABG are early stabilization of maxillary arch, obliteration of oronasal fistulas, provide bony support for the eruption of the canines and the orthodontic alignment of the teeth, support for the pyriform aperture and the nasal base enhancing the alveolus and facial appearance, to achieve better oral and dental hygiene and speech development. Secondary ABG done between the age of 9 and 11 years just before the canine eruption, i.e., when canine root is one-third to two-third complete. When the canine root is one-third to two-third complete, it will avoid the delayed risk of periodontal bone loss and root resorption. Secondary alveolar bone grafting can be done with three kinds of materials, i.e., autogenic, allogenic, and alloplastic.
The advantages of iliac autogenic bone grafts are that large quantities of cancellous bone are available and decreased operative time with two team approach. The disadvantages are the significant donor site morbidity, pain, injury to lateral cutaneous nerve, hernia, and gait disturbances. These potential complications can be overcome with the use of materials such as bone morphogenic proteins and bone morphogenic proteins with zygoma shavings.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References
- 1.Boyne PJ, Sands NR. Secondary bone grafting of residual alveolar and palatal clefts. J Oral Surg. 1972;30:87–92. [PubMed] [Google Scholar]
- 2.Steinberg B, Padwa BL, Boyne P, Kaban L. State of the art in oral and maxillofacial surgery: Treatment of maxillary hypoplasia and anterior palatal and alveolar clefts. Cleft Palate Craniofac J. 1999;36:283–91. doi: 10.1597/1545-1569_1999_036_0284_sotaio_2.3.co_2. [DOI] [PubMed] [Google Scholar]
- 3.Kortebein MJ, Nelson CL, Sadove AM. Retrospective analysis of 135 secondary alveolar cleft grafts using iliac or calvarial bone. J Oral Maxillofac Surg. 1991;49:493–8. doi: 10.1016/0278-2391(91)90172-i. [DOI] [PubMed] [Google Scholar]
- 4.Boehn A. Dental anomalies in harelip and cleft palate. Acta Odontol Scand. 1963;21(Suppl 38):1–109. [PubMed] [Google Scholar]
- 5.Björk A, Skieller V. Growth in width of the maxilla studied by the implant method. Scand J Plast Reconstr Surg. 1974;8:26–33. doi: 10.3109/02844317409084367. [DOI] [PubMed] [Google Scholar]
- 6.Waite DE, Kersten RB. Residual alveolar and palatal clefts. In: Bell WH, Proffit WR, White RP, editors. Surgical Correction of Dentofacial Deformities. Philadelphia: W.B. Saunders; 1980. pp. 1329–67. [Google Scholar]
- 7.Sweden AT. Healing of Bone Grafts: In vivo Studies of Tissue Reactions at Autografting of Bone in the Rabbit Tibia. Thesis, University of Gothenburg. 1979 [Google Scholar]
- 8.Newlands LC. Secondary alveolar bone grafting in cleft lip and palate patients. Br J Oral Maxillofac Surg. 2000;38:488–91. doi: 10.1054/bjom.2000.0300. [DOI] [PubMed] [Google Scholar]
- 9.Converse JM, Campbell RM. Bone grafts in surgery of face. Surgical Clinics of North America. 1954;34:375–401. doi: 10.1016/s0039-6109(16)34187-1. [DOI] [PubMed] [Google Scholar]
- 10.Schendel SA. Text Book of Maxillofacial Surgery. 2nd ed. Vol. 2. Philadelphia, PA: Elsevier Publications; 2002. Alveolar bone grafting. Peter Ward Booth; pp. 1062–72. [Google Scholar]
- 11.Eppley BL. Alveolar cleft bone grafting (Part I): Primary bone grafting. J Oral Maxillofac Surg. 1996;54:74–82. doi: 10.1016/s0278-2391(96)90310-9. [DOI] [PubMed] [Google Scholar]
- 12.Bilkay U, Tokat C, Ozek C, Gundogan H, Gurler T, Tegsel Z, et al. Cancellous bone grafting in alveolar cleft repair: New experience. J Craniofac Surg. 2002;13:658–63. doi: 10.1097/00001665-200209000-00012. [DOI] [PubMed] [Google Scholar]
- 13.Baqain ZH, Anabtawi M, Karaky AA, Malkawi Z. Morbidity from anterior iliac crest bone harvesting for secondary alveolar bone grafting: An outcome assessment study. J Oral Maxillofac Surg. 2009;67:570–5. doi: 10.1016/j.joms.2008.09.023. [DOI] [PubMed] [Google Scholar]
- 14.Hynes PJ, Earley MJ. Assessment of secondary alveolar bone grafting using a modification of the bergland grading system. Br J Plast Surg. 2003;56:630–6. doi: 10.1016/s0007-1226(03)00361-8. [DOI] [PubMed] [Google Scholar]
- 15.Kendrick RW, Newlands LC. Secondary alveolar clefts in the cleft lip and palate. Br J Oral Maxillofac Surg. 1998;36:220–38. doi: 10.1054/bjom.2000.0300. [DOI] [PubMed] [Google Scholar]
- 16.Weijs W.L. Seibers: early secondary closure of the alveolar clefts with mandibular symphyseal bone grafts and b- tricalcium phosphate. I journal of oral and maxillofacial surgery: 2010;39:41–429. doi: 10.1016/j.ijom.2010.02.004. [DOI] [PubMed] [Google Scholar]
- 17.Larson PE. Peterson's Principles of Oral and Maxillofacial Surgery. 2nd ed. Vol. 2. BC Decker INC publications; 2004. Reconstruction of alveolar Clefts; pp. 859–70. [Google Scholar]
- 18.Troxell JB, Fonseca RJ, Osbon DB. A retrospective study of alveolar cleft grafting. J Oral Maxillofac Surg. 1982;40:721–5. doi: 10.1016/0278-2391(82)90145-8. [DOI] [PubMed] [Google Scholar]
- 19.Kaban LB, Pogrel MA, Perott DH. 15th ed. Saunders publications; 1997. Complications in oral and Maxillofacial Surgery; pp. 289–90. [Google Scholar]
- 20.Sadove AM, Nelson CL, Eppley BL, Nguyen B. An evaluation of calvarial and iliac donor sites in alveolar cleft grafting. Cleft Palate J. 1990;27:225–8. doi: 10.1597/1545-1569(1990)027<0225:aeocai>2.3.co;2. [DOI] [PubMed] [Google Scholar]
- 21.Swan MC, Goodacre TE. Morbidity at the iliac crest donor site following bone grafting of the cleft alveolus. Br J Oral Maxillofac Surg. 2006;44:129–33. doi: 10.1016/j.bjoms.2005.04.015. [DOI] [PubMed] [Google Scholar]
- 22.Power SM, Matic DB. Gingivoperiosteoplasty following alveolar molding with a latham appliance versus secondary bone grafting: The effects on bone production and midfacial growth in patients with bilateral clefts. Plast Reconstr Surg. 2009;124:573–82. doi: 10.1097/PRS.0b013e3181addc37. [DOI] [PubMed] [Google Scholar]