Functional comparison of the dynamic hip screw and the Gamma locking nail in trochanteric hip fractures: a matched-pair study of 268 patients

Abstract

The aim of this prospective matched-pair (age, sex, fracture type, residential status, and walking ability at fracture) study was to analyse the short-term outcome after Gamma nail (GN) and dynamic hip screw (DHS) fixation, focusing especially on functional aspects (Standardised Audit of Hip Fractures in Europe [SAHFE] hip fracture follow-up forms), reoperations, and mortality. Both groups consisted of 134 patients. DHS and GN groups did not differ significantly with respect to location of residence at 4 months or returning to the prefracture dwelling (78% vs. 73%, P = 0.224). The change in walking ability at 4 months compared to prefracture situation was better in the DHS group (p = 0.042), although there was no difference in the change of use of walking aids. The frequency of reoperations during the first year was somewhat lower in the DHS group (8.2% vs. 12.7%, p = 0.318). Mortality was lower in the DHS group both at 4 months (6.0% vs. 13.4%, p = 0.061) and 12 months (14.9% vs. 23.9%, p = 0.044). Although walking ability was better and mortality lower in the DHS group, both methods are useful in the treatment of trochanteric femoral fractures.

Introduction

Internal fixation of trochanteric fractures was a significant innovation popularised by Jewett and others in 1941. It allowed early mobilisation of the patient and reduced deformity due to malunion [24]. Fixation by the original single-piece implants, however, often failed due to collapse at the fracture site and nail penetration of the head [1]. For this reason, the sliding nail plate was introduced by Pugh in 1955 [24].

During recent decades, dynamic hip screw (DHS) fixation has been thoroughly assessed, and randomised comparisons have shown it to be superior to fixed nail plates [2, 7], Enders nails [5], or the Küntscher nails [6]. The dynamic hip screw has therefore become one of the standard treatments of trochanteric fractures.

About 15 years ago, intramedullary fixation with the Gamma nail, which is based on the idea of the Y-nail proposed by Gerhard Küntscher, was introduced for the treatment of trochanteric fractures [20]. The Gamma nail consists of a dynamic sliding screw, which passes through a short intramedullary nail. The design allows sliding between the two parts to produce impaction, as in the sliding screw. The proposed theoretical advantages of the Gamma nail are reduced blood loss due to the percutaneous technique, minimal tissue damage, and shorter operation time [4, 11, 21, 23]. In addition, the possible mechanical advantage over plate fixation is that the nail is closer to the axis of weight bearing through the femoral head, and leverage is therefore reduced [4, 11, 19, 23, 24].

The use of the Gamma nail has also been found to cause many complications, the most common of which is fracture of the femoral shaft at the tip of the intramedullary nail [4, 24].

Several studies comparing the Gamma nail and DHS have shown DHS to be a better method in the treatment of trochanteric fractures [4, 23, 24], and some studies have failed to establish any significant differences between these two methods [9, 19, 25], whereas one study concluded that GN was more effective than the DHS [21]. Most of these studies have focused on fracture healing, but the functional outcome has been given less attention [23].

The aim of this study was to compare the short-term outcomes after GN and DHS fixation of trochanteric fractures, concentrating especially on the functional aspects.

Materials and methods

During the years 1991–1999, all hip fractures treated in Oulu University Hospital were prospectively registered on specific Standardised Audit of Hip Fractures in Europe (SAHFE) hip fracture follow-up forms [16, 22]. Functional parameters, i.e. location of residence, ADL functions (ability to dress and undress), walking ability, and use of walking aids were recorded at fracture (Table 2). Reoperations were recorded on a separate form. Trochanteric fractures were classified according to a modified Jensen classification as two-fragment or multi-fragment fractures [16, 17].

Table 2.

Functional parameters before the fracture and at 4 months after the fracture

	Gamma nail group					DHS group
	Before fracture		At 4 months		p-value	Before fracture		At 4 months		p-value
	n	%	n	%	(preoper.)	n	%	n	%	(4 months)
Residential status
Own home	97	72.4	67	50.0		97	72.4	75	56.0	0.349
Sheltered housing	1	0.7	1	0.7		1	0.7	0	0.0
Home for the aged	29	21.6	16	12.0		29	21.6	19	14.2
Permanent hospital inpatient	7	5.2	32	23.9		7	5.2	29	21.6
Nursing home	0	0.0	0	0.0		0	0.0	0	0.0
Acute hospital	0	0.0	0	0.0		0	0.0	3	2.2
Death	0	0.0	18	13.4		0	0.0	8	6.0
ADL (dressing)
Yes	100	74.6	60	52.6	1.000	101	75.4	77	62.1	0.151
No	34	25.4	54	47.4		33	24.6	47	37.9
Walking
Walked alone outdoors	80	59.7	24	21.1		80	59.7	36	29.0	0.175
Walked outdoors only accompanied	15	11.2	12	10.5		15	11.2	11	8.9
Walked alone indoors	34	25.4	34	29.8		34	25.4	44	35.5
Walked indoors only accompanied	5	3.7	34	29.8		5	3.7	20	16.1
Unable to walk. Able to sit	0	0.0	8	7.0		0	0.0	9	7.3
Bedbound	0	0.0	2	1.8		0	0.0	4	3.2
Walking aids
Able to walk without aids	72	54.1	8	7.0	0.572	71	53.0	13	10.5	0.113
One stick	25	18.8	19	16.7		33	24.6	26	21.0
Two sticks	5	3.8	6	5.3		6	4.5	12	9.7
One stick and tripod	0	0.0	1	0.9		1	0.7	4	3.2
Two tripods	0	0.0	0	0.0		1	0.7	0	0.0
Rollator/walking frame	31	23.4	66	57.9		21	15.7	52	41.9
Wheelchair	1	0.8	12	10.5		1	0.7	10	8.1
Does not walk	0	0.0	2	1.8		0	0.0	7	5.6
Living alone
Yes	54	40.3	34	29.8		45	33.6	30	24.2
No	80	59.7	80	70.2	0.311	89	66.4	94	75.8	0.381

There were a total of 575 trochanteric femoral fractures in 563 patients aged over 49 years. Of these, 372 were treated with GN fixation (mean age 78.6 years, range 51–96; 100 males and 272 females), and 203 were treated with DHS fixation (mean age 79.9 years, range 51–101; 50 males and 153 females). The patients treated with GN fixation were cross-matched with the patients treated with DHS fixation for age, sex, location of residence at fracture, walking ability at fracture, and fracture type. Cross-matching was performed by a statistician, and 134 pairs were found. The cross-matching data are presented in Tables 1 and 2.

Table 1.

Demographic data

	Gamma nail group		DHS group		p-value for difference
	n	%	n	%
Sex
Male	23	17	23	17
Female	111	83	111	83
Age on operation day
Mean (range)	80 (52–93)		80 (51–94)		0.994
Marital status
Unmarried	16	11.9	20	14.9	0.516
Married	28	20.9	32	23.9
Divorced	4	3.0	7	5.2
Widow/er	86	64.2	75	56.0
Lived alone at the time of fracture
No	80	59.7	89	66.4	0.311
Yes	54	40.3	45	33.6
Need for home help at the time of fracture
Mean h/week		1.5		1.8
Minimum		0.0		0.0	0.356
Maximum		21.0		30.0
Side of fracture					0.714
Right	66	49.3	70	52.2
Left	68	50.7	64	47.8
	134	100.0	134	100.0
Type of fracture
Trochanteric two-fragment fracture	54	40.3	54	40.3
Trochanteric multifragment fracture	80	59.7	80	59.7

Follow-up was continued for 4 months by recording the same functional parameters that were recorded on admission. The changes in residential status and walking ability with respect to the situation at fracture were also considered (worse, same, better) (Table 3). Mortality and reoperation rate were recorded for up to 1 year after the fracture.

Table 3.

Residential status, dressing and undressing, walking-ability, use of walking aids, and mortality evaluated at 4 months as the change in comparison to the situation before the fracture

	Gamma nail group		DHS group		p-value for difference
	n	%	n	%
Residential status
Worse	29	25.0	28	22.2	0.224
Same	85	73.3	98	77.8
Better	2	1.7	0
ADL (dressing)
Worse	29	25.7	24	19.7	0.401
Same	81	71.7	92	75.4
Better	3	2.7	6	4.9
Walking
Worse	78	69.0	70	56.9	0.042
Same	34	30.1	46	37.4
Better	1	0.9	7	5.7
Walking aids					0.093
Worse	79	69.9	90	73.3
Same	32	28.3	25	20.3
Better	2	1.8	8	6.4
Mortality 4 months					0.061
Alive	116	86.6	126	94.0
Died	18	13.4	8	6.0
Mortality 1 year
Alive	102	76	114	85
Died	32	24	20	15	0.044

Operative techniques

All operations were on a traction table under spinal anaesthesia. Low-molecular heparin as thromboembolic prophylaxis and preoperative single-shot antibiotic prophylaxis (1.5-g cefuroxime i.v.) were used in all cases. Dynamic hip screws and Gamma nails were inserted by using the previously presented standard techniques in line with the recommendations of the manufacturers of the Gamma nail (Howmedica Ltd., London, UK) and the DHS implant (Stratec Medical, Oberdorf, Switzerland). For both implants, we tried to get the sliding screw into the lowest third of the neck in A-P projection, into the middle part of the neck in latereral projection, and as close as 5 mm to the subchondral bone, as recommended in earlier studies [19]. The selection between these two methods was done by the surgeon according to his preference. The patients were mobilised after an X-ray examination on the first postoperative day in both groups.

Statistical analysis

Statistical analysis was performed by a statistician. The material was processed and analysed using the SPSS for Windows 12.0.1 software (SPSS Inc., USA). Chi-square test with Yates correction, Fischer’s exact test, and nonparametric Mann-Whitney-U test were used to evaluate the significance of the differences. A difference was considered to be statistically significant when p < 0.05.

Results

Delay of operation and hospital stay

DHS and GN groups did not differ significantly with respect to delay from admission to operation (1.2 days vs. 1.3 days, p = 0.64) and length of hospital stay (8.8 days vs. 7.1 days, p = 0.28).

Residential status

DHS and GN groups also did not differ significantly with respect to living in their own homes at 4 months (56% vs. 50%, p = 0.35, Table 2) or returning to the prefracture dwelling (78% vs. 73%, p = 0.224, Table 3).

Walking ability, use of walking aids, dressing and undressing

DHS and GN groups did not differ with respect to walking ability at 4 months (p = 0.18, Table 2). However, the change in walking ability at 4 months compared to prefracture situation was significantly better in the DHS group (p = 0.042 ), although there was no significant difference in the change of use of walking aids (p = 0.09, Table 3). The groups also did not differ in the ability to dress and undress at 4 months (p = 0.40).

Pain and use of analgesics

In the GN group, pain on weight bearing was somewhat more common (p = 0.108), as well as consumption of analgetics (p = 0.111).

Mortality

In the DHS group, mortality was lower both at 4 months (6.0% vs. 13.4%, p = 0.061) and 12 months (14.9% vs. 23.9%, p = 0.044, Table 3).

Reoperations

The frequency of reoperations within 1 year was somewhat lower in the DHS group (11 vs. 17, p = 0.318). Twenty five of 28 reoperations occurred within the first 4 months, and the reasons (DHS group/GN group) included: 11 fracture displacements (4/7), six loss of position of the nail (2/4), three infections (2/1), three femoral head collapses (1/2), and two refractures (0/2). The reoperation types were revision of fixation in 14 of the cases, arthroplasty in five, Girdlestone in one, and other procedures in the remaining four. Three reoperations were performed after the first 4 months and included one removal of osteosynthesis in the GN group and one arthroplasty and one revision of fixation in the DHS group. Nonunion was the reason for two reoperations.

Discussion

The incidence of hip fractures has been predicted to increase because of the ageing of the population [18, 19]. It is thus important to strive to improve treatment and to develop better surgical devices, of which the introduction of the Gamma nail is one example. The new treatment modalities must prove their usefulness and superiority over the old methods, even in terms of functional outcome, which is an aspect lacking from most of the earlier comparisons between Gamma nails and other fixation materials.

Our study was a prospective matched-pair study comparing gamma nail fixation and DHS fixation in trochanteric fractures by focusing on functional recovery, reoperation rate, and mortality. Functional outcome was assessed at 4 months, by which time ADL, walking ability, and household activities have been shown to have reached a constant level [3, 12].

DHS fixation required a slightly but not significantly longer hospital stay than GN fixation. This has been stated in many earlier reports [4, 19, 25]. The average lengths of hospital stay in this study were definitely shorter in both groups than those reported in the meta-analysis of 10 trials on Gamma versus DHS nailing by Parker and Pryor in 1996, where the means ranged from 12 to 39 days for DHS and from 12 to 37 days for Gamma nail [23]. The explanation for this difference is the Finnish system of elderly care; the catchment area of our hospital has an extensive network of health centre hospitals capable of providing rehabilitation at a very early postoperative phase.

The facts that the GN is said to be more rigid and to allow full weight bearing earlier than the DHS even in cases of very complex fractures, and that DHS fixation requires more extensive surgery than GN fixation did not have any marked effects on the functional outcome. On the contrary, walking ability evaluated as the change compared to the preoperative situation was better in the DHS group. Similar results have also been reported by Hoffman and Lynskey [13], whereas Goldhagen et al. [9], Sabharwal et al. [25], and Park et al. [21] observed no significant differences. The greater impairment of walking ability after GN fixation than after DHS fixation in this study may be at least partly due to the higher rate of complications in the GN group.

The cutting-out from the femoral head is regarded as a typical complication of DHS fixation [4, 10, 19], but in our study such cutting-out occurred more often with gamma nails. Our findings are in agreement with some earlier reports [9, 14, 25], whereas Hoffman and Lynskey [14], Fornander et al. [8], and Park et al. [21] reported roughly similar cutting-out rates for both methods. We agree with Bridle et al. [4] in that cutting-out from the femoral head is usually due to a technical error rather than implant dysfunction.

Complications specific to GN are fractures around the greater trochanter, fracture displacement by nail insertion, and fractures of the shaft of the femur [21]. Fracture of the femoral shaft at the distal end of the intramedullary part of the implant is considered the most serious complication of GN fixation [4, 19, 23, 24]. Both of the two fractures of the femoral shaft that occurred in our GN group were associated with a fall. The fracture rate of the femoral shaft, which was 1.5% in our study, was similar to those published in earlier reports [11, 20].

Most studies have failed to reveal any difference in postoperative mortality between GN and DHS fixation [10, 14, 19, 24]. Some studies, however, have reported higher mortality after GN fixation [13, 25] and some after DHS fixation [4, 8]. Our findings of higher mortality after GN fixation are in line with the earlier results, and we think that the difference was at least partly due to the higher number of complications requiring reoperations associated with the GN fixation.

It is concluded that both methods are useful in the treatment of trochanteric femoral fractures, although the results were slightly in favour of DHS fixation with respect to walking ability and mortality.