Comparison of the 2-Incision and Mini-Incision Posterior Total Hip Arthroplasty Technique: A Retrospective Match-Pair Controlled Study

Procedures 

Patients aged 18 years or older at the time of total hip arthroplasty, and who were operated on between December 2002 and June 2004, had either a 2-incision or posterior mini-incision operation. Two hundred eighty-four operations were performed on patients between 18 and 75 years of age who were eligible for the matched-pair analysis. Patients had the diagnosis of osteoarthritis, rheumatoid arthritis, posttraumatic arthritis, or a displaced femoral neck fracture. More than 90% of the patients in this initial cohort had the diagnosis of osteoarthritis. All patients who were operated were candidates for total hip arthroplasty and had obtained medical clearance for the operation. The first minimally invasive surgery (MIS 2)-incision technique was performed at our center on April 10, 2001. The mini-posterior and 2-incision techniques have been used for approximately 5 years at our center. Approximately 200 two-incision cases and 300 mini-posterior cases had been performed before this match-pair cohort. These cohorts of patients were not considered part of the initial “learning curve.” All patients in the match-pair cohort had the diagnosis of osteoarthritis. The only patients who had a more traditional total hip arthroplasty during this time had hardware inserted from prior hip surgery (3 cases) or had morbid obesity (4 cases). The pairs were matched by the criteria listed in Table 1. All operations were performed by a single surgeon.

Table 1. Baseline and Operative Information

	Variable	Two-Incision Matched Patients	Mini-Incision Matched Patients
	Age
	N	43	43
	Mean	57.4	59.1
	SD	6.3	8.2
	(Min, Max)	(40, 68)	(43, 74)
	Gender, n (%)
	Male	24 (54)	24 (54)
	Female	19 (44)	19 (44)
	Body mass index category, n (%)	43	43
	Normal (<25 kg/m^2)	21 (48%)	21 (48%)
	Overweight (25-30 kg/m2)	15 (35%)	15 (35%)
	Obese (>30 kg/m2)	7 (16%)	7 (16%)
	Preoperative diagnosis, n (%)
	Osteoarthritis	43 (100)	43 (100)
	Other	0 (0)	0 (0)
	Length of follow-up
	N	43	43
	Mean	1.2	1.1
	SD	0.17	0.17
	(Min, Max)	(0.5, 2.0)	(0.5, 2.0)
	WOMAC pain
	N	43	43
	Mean	9.1	9.0
	SD	2.8	2.9
	(Min, Max)	(4, 15)	(2, 14)
	WOMAC physical function
	N	43	43
	Mean	28.4	29.8
	SD	9.6	9.8
	(Min, Max)	(5, 45)	(8, 50)
	WOMAC stiffness
	N	43	43
	Mean	3.9	4.4
	SD	1.6	1.4
	(Min, Max)	(1, 8)	(2, 8)

Our 43 match-pair series was selected from a pool of 284 patients. This group was the cohort selected because they had completed the weekly patient diaries, which was so important for evaluating early postoperative follow-up and rehabilitation. The mini-posterior and 2-incision techniques have been used for approximately 5 years at our center. Clearly, some patients came to the author for the 2-incision technique and therefore were somewhat self-selected because they wanted this procedure. However, patients who did not meet the author's criteria, which has been well defined, did not receive the 2-incision procedure. There were no selection criteria for the mini-posterior incision; thus, any patient who did not fit the criteria for the 2-incision procedure, whether those patients wanted the 2-incision or not, were included in the mini-posterior group. Patients who were in the 2-incision group had selected criteria of age less than 70 years and body mass index less than 40, do not have a Crowe dysplasia classification of class III or IV, and do not have an excessive anterior-posterior bow of the femur, which would not allow safe use of a straight fully coated femoral stem. In addition, the diameter of the intramedullary canal of the femur could not exceed 18 mm, and the size of the acetabular cup could not exceed 64 mm.

There were no restrictive criteria for the posterior mini-incision patients. Preoperatively, there was only one difference in the preoperative education class that patients attended. In the class for the posterior mini-incision patients, the only difference was instruction in dislocation precautions, whereas the 2-incision patients were not restricted by dislocation precautions. No patient in either group had an abduction pillow. The posterior group was instructed in avoidance of extreme flexion and internal rotation. This is the only difference in hip precautions between the 2 groups. Patients were educated on the benefits of early mobility in preventing deep venous thrombosis and the benefits of early discharge in preventing nosocomial infection. Of the 43 patients with the posterior mini-incision and of the 43 patients with the 2-incision operations, 40 and 25 patients, respectively, attended preoperative class. Most patients attended the preoperative class. The only patients who did not attend the preoperative class were nonlocal patients, and they were seen immediately after my preoperative visit, which was usually the day before the procedure. All patients in this category received the same education from our orthopedic nurse practitioner. All patients were instructed to maintain a postoperative diary of their recovery.

In the hospital, all patients were treated with the same clinical hip pathway. Anesthesia was a short-acting spinal anesthetic using hyperbaric 0.75% Bupivacaine (Abbott) adjusted for the height and weight of the patient. The pain management included administration of valdecoxib, 40 mg (Pharmacia and Upjohn); oxycodone SR, 10 to 20 mg (Roxane); and ondansetron, 4 mg (GlaxoSmithKline). In the recovery room, 25 to 50 μg of fentanyl (Abbott) every 5 to 10 minutes was given as needed. Once the patients had returned to their room, only oral medication was used. Deep venous thrombosis prevention included the short-acting spinal anesthetic, rapid mobilization of the patient, calf high compression stockings, calf pneumatic compression, and 325 mg of aspirin each day in the hospital and for 3 months postoperatively.

Patients were discharged home as soon as they were judged safe by physical therapy and internal medicine, and this could be as early as the day of the surgery. Following discharge, patients were encouraged to use aspirin and plain Tylenol for pain medications and to avoid oral narcotics. All patients were instructed to use a regular pillow between their legs while sleeping for 6 weeks for comfort. Neither group used an abduction pillow. They were told they were allowed to drive their car when they were off narcotic pain medication and had appropriate muscle control of their leg. All patients were allowed to achieve weight-bearing as tolerated. Both patient groups were treated equally with respect to weight-bearing as tolerated without regard to stem type. The type of femoral stem did not inhibit or prevent patients from being able to ambulate immediately. A tight press fit was achieved at the time of surgery with either stem to allow full immediate weight-bearing.

Surgical Technique 

The surgical technique for the 2-incision operation has been described, [1], [11], [12], [13] as has the technique for the posterior mini-incision [1], [6], [8], [16], [17]. The implants used were a Trilogy acetabular component for all patients (Zimmer, Warsaw, Ind). We used 1 or 2 screws to augment the fixation of all acetabular components. All patients with the 2-incision total hip arthroplasty had a fully coated VerSys femoral stem (Zimmer). All surgeons reporting cases in the initial Journal for Bone and Joint Surgery (JBJS) American Orthopaedic Association (AOA) symposium used the full coat stem [11]. This implant is straight and acts more like a femoral nail, allowing easier insertion and better fit through the small posterior incision. The VerSys FullCoat stem was implanted in 11 (26%) of 43 patients with the posterior mini-incision technique, whereas 32 (74%) of 43 had a VerSys Fiber Metal MidCoat stem (Zimmer). The full coat stem was used in these cases in a Dorr C type femur or when proximal fill or bone quality was thought to be an issue.

Clinical Assessment 

Demographics included sex, primary diagnosis, prior operations to the hip, height, weight, range of motion of the hip to be operated, leg length discrepancy, and the presence or absence of a Trendelenburg gait pattern. Clinical assessments were collected preoperatively, and postoperatively at 3 months, 6 months, and 1 year. The data for this study were extracted and summarized on May 25, 2005. At that time, there were 15 of 43 cases in each group with 1-year follow-up data available. This data included 1-year radiographs. The instruments used were the (1) Harris Hip Score, (2) the Medical Outcomes Study 36-Item Short-Form Health Survey (MOS SF-36), (3) the Medical Outcomes Study 6-Item Sleep Scale, and (4) the Western Ontario and McMaster (WOMAC) Osteoarthritis Index. Before surgery, the hospital admission date (and time) was recorded. Data recorded during surgery included the surgical technique, operating room time, fluoroscopy time, blood loss measured from the cell saver, transfusion of blood and blood products, and intraoperative complications.

The MOS SF-36 is a 36-item generic survey that collects information on 8 health concepts: physical functioning, role-physical, bodily pain, general health, vitality, social functioning, role-emotional, mental health, and reported health transition. Individual items were scored on an integer scale, and concept scores were generated from composites of the item scores. Composite scores ranged from 0 to 100, with a score of 100 being the best. In addition, composite mental and physical health scores ranging from 0 to 100 were constructed using population norms.

The WOMAC Osteoarthritis Index is a 24-item disease-specific survey that collects information on 3 areas of patient symptoms: pain, stiffness, and physical function. Individual items were scored on an integer scale, and item scores were combined to generate each area. In addition, a total WOMAC score based on all 24 items was generated.

The MOS Sleep Scale is a 6-item quality of life survey, which assesses patient sleep patterns.

The Harris Hip Scale is a preoperative and postoperative assessment of pain, function, range of motion, and absence of deformity designed to be equally applicable to different hip problems and different methods of treatment [18].

After discharge, for the first 6 weeks, all patients completed a patient diary recording the time they returned to activities of daily living without assistance (bathing, dressing, getting in and out of bed and chair), driving, work, and the discard of an assistive device for ambulation.

Complication Assessment 

Complications evaluated in this study included leg length discrepancy, component malposition, subsidence, periprosthetic fracture, sciatic or femoral nerve injury, lateral femoral cutaneous nerve injury, infection, and wound problems.

Radiographic Assessment 

Radiographs were performed immediately postoperative, at 6 weeks, 3 months, 6 months, and annual visits. These were an anterior-posterior pelvic radiograph that included the acetabulum and femoral stem as well as a lateral (cross-table) of the femur and acetabulum. For this study, the acetabular inclination and anteversion were measured from the 6-week or 3-month radiograph. From the radiograph with the longest follow-up, any evidence of component position change (migration) or radiolucent line formation, osteolysis, and heterotopic ossification was measured. The radiographic measurements were performed on the mini-posterior and 2-incision patients at the longest follow-up period. There were 15 of 43 patients with 1-year follow-up in each group. All patients in the 2-incision group had intraoperative fluoroscopy to confirm cup and stem insertion.

Statistical Methods 

Forty-three matched pairs provided at least 80% power at a 2-sided α level of .05 to detect clinically meaningful differences in study variables Type I error rates (α level, .05) were adjusted for multiplicity using a Bonferroni correction [19]. For comparisons made only once (eg, comparisons of preoperative, operative, immediate postoperative variables, return to function, and complication rates), statistical significance was inferred at the .05 level. Unmatched analyses were performed on the matched data using SAS (v 9.13) for the Microsoft Windows 2000 platform.

Clinical Assessment 

Fig. 1 presents mean scores by surgical technique at each time of assessment. Benefits in both techniques are confirmed by the large positive gains (improvement) in Harris Hip Scores, MOS SF-36 Physical Health and Physical Function scores, as well as large negative gains (improvement) in WOMAC pain, stiffness, and physical function seen with increasing time since total hip arthroplasty.

View Large Image Download to PowerPoint

Fig. 1. Average Harris Hip Score and selected SF-36 and WOMAC scores at each time of assessment.

Patients were considered compliant if they completed the patient diary entries during postoperative weeks 1 through 6 or if a scheduled clinical examination was attended at postoperative weeks 2 and 6, months 3 and 6, and at 1 year. Two-incision patients had a mean compliance of 91% (range, 86%-95%) during the first 6 postoperative weeks, whereas posterior mini-incision patients had a mean of 88% (range, 77%-95%) (P = .68). At 6 weeks postoperatively, the compliance was 41 (95.3%) of 43 patients in both groups; at 6 months, 33 (76.7%) of 43 two-incision patients and 28 (65.1%) of 43 posterior mini-incision patients (P = .23); and at 12 months, 26 (70.3%) of 37 two-incision patients and 26 (86.7%) of 30 posterior mini-incision patients (P = .09). At 12 months, not all patients had attained their 1-year postoperative check-up.

The postoperative clinical outcomes, which had statistically significant differences, are displayed in Fig. 2. There were 2 items in the Harris Hip Score that favored the 2-incision patients, and these were no assistive device needed at 2 weeks postoperatively (P = .001) and negotiating stairs without a railing at 5 weeks postoperatively (P = .004). There were 2 WOMAC pain subscale items that favored the posterior mini-incision, and these were no pain at night at 6 weeks (P = .0028) and pain at night in bed at 6 weeks (P = .004). At 6 weeks postoperatively, the total WOMAC pain subscale score for the posterior mini-incision patients was a mean of 1.22 (range, 0-10), and for the 2-incision group was a mean of 3.31 (range, 0-11) (P = .003). Prescription antiinflammatory use had significantly greater prevalence in 2-incision patients (47%) versus mini-incision (23%) (P = .04). At 6 weeks postoperatively, there were 2 MOS Sleep Scale items that favored the posterior mini-incision patients, which were no trouble falling asleep (P = .0004), and if awakened, no trouble falling back to sleep (P = .003). We have looked closely at the pain scores and graphs, and we cannot say with certainty that there is a difference in looking at our aggregate data. The numerous pain graphs at different points give data to support either group being beneficial in regards to pain relief but at different times.

View Large Image Download to PowerPoint

Fig. 2. Statistically significant patient outcomes at each time of assessment.

Results, which measured return to function, favored the 2-incision patients for 8 of the 9 items. Two of these items had statistical difference: the time to resume driving, which was 13 days (range, 2-31) for 2-incision patients and a mean of 24 days (range, 6-32) for posterior mini-incision patients (P = .04); and the time to resume shopping, which had a mean of 14 days (range, 3-24) for 2-incision patients and a mean of 26 days (range, 6-37) for mini-incision patients (P = .01). Length of hospital stay was significantly shorter (<0.0001) for 2-incision patients with a mean of 30.7 hours (range, 12.9-55.7) versus mini-incision patients with a mean of 44.6 hours (range, 13-102). There were only 3 patients in the mini-posterior group who stayed longer than day 2. Two patients were discharged on postoperative day 3. None of these outliers in the mini-posterior group were related to any orthopedic problems. One patient aged 58 had dizziness and had difficulties being cleared by physical therapy. Another 68-year-old female patient had diabetes that was difficult to manage postoperatively and experienced lightheadedness and extreme calf spasms that significantly delayed mobilization. One patient was discharged on postoperative day 4. A 73-year-old woman was unable to rehabilitate quickly secondary to multiple medical problems and required 2 units of packed red blood cells (PRBCs). She also had a nutritional consult. None of these outliers in the mini-posterior group were related to any orthopedic problems.

Data from the operation that has statistical difference was blood loss (as collected in a cell saver), which had a mean of 366 ± 215 mL (range, 150-1400 mL) for the 2-incision group and a mean of 247 ± 90 mL (range, 100-450 mL) for the posterior mini-incision patients (P = .001). Three 2-incision patients and one posterior mini-incision patient received allogenic blood transfusions. Although we had 1 outlier for blood loss in the 2-incision total hip arthroplasty group, there were no adverse encounters with early discharge in either group. In fact, the patient with an estimated blood loss (EBL) of 1400 mL was discharged the next day after receiving 1 unit of blood. This 65-year-old woman completed all of our discharge requirements and was discharged on the following day. Operative time for the 2-incision patients was a mean 93.7 ± 90 minutes (range, 72-135 minutes) as compared to a mean 61.7 ± 60 minutes (range, 39-111 minutes) for posterior mini-incision patients (P = .002). There were no complaints of leg length discrepancy, and only 3 patients had a measured length difference (<5 mm) from the preoperative evaluation.

Radiographic Results 

The mean cup abduction angle for the 2-incision hips was 49.4° ± 4.2° (range, 41°-60°), and for the posterior mini-incision hips was 45.6° ± 5.3° (35°-57°) (P = .0003). Our target number was 45° for both approaches. We had 4 outliers in the 2-incision group that were greater than 50° and 2 in the mini-posterior group. Most of the 2-incision cup abduction angles were between 45° and 50°. The cup anteversion for the 2-incision hips had a mean 20.2° ± 5.3° (range, 5°-25°), and for the posterior mini-incision hips was 18.4° ± 7.9° (range, 0°-35°) (P = .2624). The stem was implanted in 5° varus in 1 hip in the posterior mini-incision group, and in one 2-incision patient, the stem was in 3° varus. All other stems were in neutral alignment.

Complications 

There were 2 (4.6%) of 43 hips with complications in the 2-incision group and none in the posterior mini-incision group. The 2 complications were 1 intraoperative femoral neck fracture, which was treated with a cerclage wire, and 1 suture abscess, which was treated with irrigation and debridement and antibiotics. For the 2-incision patients, 6 of 43 complained of some numbness in the anterolateral thigh, but none had frank myalgia paresthetica symptoms. Of 43 patients, 3 have permanent numbness in the distribution of the lateral femoral cutaneous nerve, the other 3 resolved completely. No patients complained of nerve pain. All patients selected for the 2-incision surgical technique are told of this possibility preoperatively because this is not a complication with the mini-posterior approach. Leg length assessment at 6 weeks revealed that 33 (85%) 2-incision and 32 (84%) mini-posterior incision patients had equal leg lengths. There were 6 (15%) patients in the 2-incision group with leg length discrepancy at 6 weeks. There were 6 (16%) patients in the mini-posterior group at 6 weeks with leg length discrepancy. Of the 20 cases per group with leg length assessment available at 6 months, both groups were 100% equal. No patient had a leg length discrepancy greater than one fourth of an inch. There were no sciatic, peroneal, or common femoral nerve injuries. There were no components that subsided. No patient in this series required revision.