Internal fixation compared with total hip
replacement for displaced femoral neck
fractures in the elderly
A RANDOMISED, CONTROLLED TRIAL
J. Tidermark, S. Ponzer, O. Svensson, A. Söderqvist, H. Törnkvist
From Stockholm Söder Hospital and Umeå University Hospital, Sweden
he treatment algorithms for displaced fractures of
T the femoral neck need to be improved if we are to
reduce the need for secondary surgery. We have studied
102 patients of mean age 80 years, with an acute
displaced fracture of the femoral neck. They were
randomly placed into two groups, treated either by
internal fixation (IF) with two cannulated screws or total
hip replacement (THR). None showed severe cognitive
dysfunction, all were able to walk independently, and all
lived in their own home. They were reviewed at four, 12
and 24 months after surgery. Outcome measurements
included hip complications, revision surgery, hip
function according to Charnley and the health-related
quality of life (HRQoL) according to EuroQol (EQ-5D).
The failure rate after 24 months was higher in the IF
group than in the THR group with regard to hip
complications (36% and 4%, respectively; p < 0.001),
and the number of revision procedures (42% and 4%, p
< 0.001). Hip function was significantly better in the
THR group at all follow-up reviews regarding pain (p <
0.005), movement (p < 0.05 except at 4 months) and
walking (p < 0.05). The reduction in HRQoL (EQ-5D
index score) was also significantly lower in the THR
group than in the IF group, comparing the pre-fracture
situation with that at all follow-up reviews (p < 0.05).
The results of our study strongly suggest that THR
provides a better outcome than IF for elderly, relatively
healthy, lucid patients with a displaced fracture of the
femoral neck.
J Bone Joint Surg [Br] 2003;85-B:380-8.
Received 10 June 2002; Accepted after revision 28 November 2002
J. Tidermark, MD, PhD, Consultant Orthopaedic Surgeon
S. Ponzer, MD, PhD, Consultant Orthopaedic Surgeon
A. Söderqvist, RN
H. Törnkvist, MD, PhD, Consultant Orthopaedic Surgeon
Department of Orthopaedics, Karolinska Institute, Stockholm Söder Hospi-
tal, S-118 83 Stockholm, Sweden.
O. Svensson, MD, Professor
Department of Surgery and Perioperative Sciences, Umeå University Hos-
pital, S-90185 Umeå, Sweden.
Correspondence should be sent to Dr J. Tidermark.
©2003 British Editorial Society of Bone and Joint Surgery
doi.10.1302/0301-620X.85B3.13609 $2.00
380
The incidence of hip fractures is increasing throughout the
world. The annual number is estimated to rise from 1.7 mil-
lion in 1990 to 6.3 million by the year 2050.1 This will pro-
vide a major challenge for every health-care system.
Treatment for the displaced fracture2 of the femoral neck in
the elderly is still controversial and in a large meta-analysis
the complication rate is reported to be 49%.3 Early redis-
placement and nonunion occurred in 33% and avascular
necrosis (AVN) in 16% of patients. Second operations were
required in 35%. Recent data indicate that the outcome
regarding the health-related quality of life (HRQoL), after
an uneventfully healed displaced fracture of the femoral
neck, is significantly impaired compared with before the
fracture and also compared with an uneventfully healed
undisplaced fracture.2,4,5
Total hip replacement (THR) is established in osteo-
arthritis and rheumatoid arthritis and has also been used as a
primary procedure for displaced fractures of the femoral
neck.6-11 The major drawback has been the fear of disloca-
tion, reported to occur in 11% (0 to 18).3 There have been
few randomised controlled trials which have compared
internal fixation (IF) with THR.6-11 Despite a relatively high
dislocation rate, particularly in patients with cognitive
impairment,8 THR is recommended as the primary treat-
ment because of the low rate of reoperation and better
immediate function of the hip. None of these studies, how-
ever, evaluated the HRQoL. The population of patients with
fractures of the femoral neck is heterogeneous and there is a
need to establish management criteria and guidelines.3 We
present the outcomes after displaced fractures of the femo-
ral neck in elderly patients treated either by IF or THR.
Patients and Methods
We have studied 87 women (79%) and 23 men with a mean
age of 80 years (70 to 96) who sustained acute, displaced
fractures of the neck of the femur (Garden III and IV).2 The
inclusion criteria were age ≥70 years, no evidence of severe
cognitive dysfunction, giving 3 or more correct answers on
a ten-item Short Portable Mental Status Questionnaire
(SPMSQ),12 domestic independence (i.e. not institutional-
ised) and ability to walk with or without walking aids.
Patients with fractures not suitable for internal fixation, i.e.
pathological fractures, displaced fractures more than 24
THE JOURNAL OF BONE AND JOINT SURGERY
 INTERNAL FIXATION COMPARED WITH TOTAL HIP REPLACEMENT FOR DISPLACED FEMORAL NECK FRACTURES IN THE ELDERLY 381

Randomised patients
110

Randomised THR Randomised IF

55 55

Excluded, 6 Excluded, 2
-see text -see text

Included & received THR Included & received IF

49 53

Deceased, 5 Lost, 1 Deceased, 10 Lost, 2

-none re-operated -revised due to -hip complications 3 -none re-operated
dislocation

THR IF
24-month follow-up 24-month follow-up
43 41

Hip complication, 1 Hip complications, 16

-periprosthetic # -non-unions, 10
-AVN, 6

Re-operated Re-operated, 14
-Open reduction and IF -arthroplasties, 10
of periprosthetic # -removal of screws, 4

THR IF
No hip complications No hip complications
42 25

Fig. 1

The surgical outcome for the 102 patients included in the study.

hours old, and patients with chronic arthritis, either rheuma-
toid or osteoarthritis, were not included. After acceptance
by an anaesthetist, the patients were randomly allocated
(sealed-envelope technique) to be treated either by IF with
two cannulated screws or by primary THR.
The study was performed according to the Helsinki dec-
laration and the protocol was approved by the local Ethics
Committee.
In the THR group two patients with aortic valve stenosis
were considered to be unfit for surgery and were excluded.
One patient developed a urinary infection while awaiting
surgery and two patients changed their mind after randomi-
sation and opted out of the study. In the IF group one patient
changed her mind after randomisation and refused to partic-
ipate and two patients, one in each group, were excluded
after treatment when rheumatoid arthritis was diagnosed,
although it did not affect the hip. Therefore, 102 patients
remained, 53 patients in the IF group and 49 in the THR
group (Fig. 1).
Operative techniques. One of two surgeons (JT or HT)
carried out the primary operations in both groups. Both
were general orthopaedic surgeons specialising in trauma
and experienced in both the procedures used in this study.
With the patient on a fracture table, the fractures were
reduced by closed manipulation, with the help of an image
intensifier, to neutral or slightly valgus impaction, and inter-
VOL. 85-B, No. 3, APRIL 2003
nally fixed with two cannulated screws (Olmed; De Puy,
Sweden). The screws were positioned according to the
recommendations of Lindequist and Törnkvist.13 We cate-
gorised the reductions as good (displacement <2 mm,
Garden angle 160˚ to 175˚, posterior angulation <10˚), fair
(displacement <5 mm, Garden angle 160˚ to 175˚, posterior
angulation <20˚) or poor (displacement >5 mm, Garden
angle <160˚ or >175˚, posterior angulation >20˚). The posi-
tion of the screw was good if the tips of the screws were
within 5 mm of the subchondral bone. In the anteroposterior
(AP) projection the distal screw entered the lateral cortex at
the level of the lesser trochanter and lay on the calcar femo-
rale. The proximal screw was introduced at least 2 cm above
and parallel to the distal one (the angle being within 10˚ of
the distal screw). In the lateral projection, the screws were
placed parallel to each other in the central or posterior third
of the femoral head and neck. All patients were given low-
molecular-weight heparin (Fragmin; Pharmacia, Sweden)
preoperatively and daily for approximately ten days after
surgery. No antibiotic prophylaxis was given to patients in
this group.
For THR the surgeons used an anterolateral approach
(modified Hardinge14), with the patient in the lateral decu-
bitus position. The Exeter modular stem (Stryker, Sweden)
with a head diameter of 28 mm, and the OGEE acetabular
component (De Puy, Sweden) were used. All patients were
382 J. TIDERMARK, S. PONZER, O. SVENSSON, A. SÖDERQVIST, H. TÖRNKVIST

Table I. The preoperative details of 102 elderly patients with displaced fractures of the femoral
neck randomised to either IF or THR; there were no significant differences between the groups
THR IF
Number of patients 49 53
Mean (± SD) age in years 79.2 ± 5.0 81.4 ± 6.6
Mean (± SD) cognitive function SPMSQ 9.0 ± 1.1 8.7 ± 1.6
Mean (± SD) EQ-5D index score prefracture 0.80 ± 0.22 0.84 ± 0.13
Gender (%)
F 40 (82) 42 (79)
M 9 11
Number with mobility with no walking aid or just one stick (%) 45 (92) 46 (87)
Number with ADL with index A or B (%) 48 (98) 51 (96)
Number with co-morbidity A or B (%) 40 (82) 44 (83)

Table II. Hip function according to the Charnley score for the 95 patients available at the four-month follow-up, the
92 available at 12 months and the 84 available at 24 months (1 = total disability, 6 = normal state)
THR IF
Percentage of Percentage of
Follow-up patients with patients with
(months) Mean value scores of 5 and 6 Mean value scores of 5 and 6 p value
Pain 4 5.7 85 4.7 55 <0.001
12 5.3 76 4.5 50 <0.005
24 5.6 84 4.7 54 <0.005
Movement 4 5.0 65 4.7 49 NS
12 5.0 70 4.6 44 <0.01
24 4.9 67 4.5 46 <0.05
Walking 4 4.3 33 3.5 15 <0.001
12 4.6 50 3.9 24 <0.01
24 4.5 51 3.8 32 <0.05

given low-molecular-weight heparin (Fragmin) preopera-
tively and daily for approximately ten days after surgery.
Cefuroxim (Zinacef, 1.5 g; GlaxoSmithKline, Sweden) was
given preoperatively followed by two additional doses
during the first 24 hours. We categorised the position of the
stem as good if it was in the neutral position or slight
valgus (<5˚ of varus). The cup was considered to be in a
good position if the lateral opening in the AP view was 30˚
to 50˚ from the horizontal and if the anteversion on the lat-
eral view was less than 30˚. Leg length was also assessed
on the radiographs. The operating time, the intraoperative
blood loss and the need for blood transfusion were all
recorded.
Patients in both groups were mobilised bearing full
weight as tolerated. We informed patients in the THR group
about mobilisation techniques and allowed them to sit on a
high chair immediately after surgery. They abandoned their
crutches at their own convenience. After six weeks there
were no restrictions.
Primary assessment and follow-up. The primary assess-
ment included confirmation of the inclusion and exclusion
criteria and any comorbidity.15 The patients were inter-
viewed about their mobility, activities of daily living (ADL)
and HRQoL according to the EuroQol (EQ-5D) during the
last week before the fracture (Table I).16,17 They attended
for clinical and radiological review at four (mean 4.3, SD
0.6), 12 (mean 12.4, SD 1.3) and 24 months (mean 24.2, SD
0.9). We assessed hip function according to Charnley’s
numerical classification18 and asked the patients to rate their
HRQoL according to the EQ-5D. An unbiased observer (a
research nurse, not involved in the surgery or clinical deci-
sions) assessed all clinical variables except movement of the
hip. The research nurse could not be blinded as to the type
of surgical intervention.
We graded comorbidity as (A) full health, (B) another ill-
ness not affecting rehabilitation and (C) another illness
which affected rehabilitation.15 The Katz ADL index status
is based on an evaluation of the patient’s functional inde-
pendence or dependence on others for bathing, dressing,
feeding, going to the toilet, transferring and continence.16
An ADL index of A indicates independence in all six func-
tions and index B, independence in all but one of the six
functions. Indices C to G indicate dependence in bathing
and at least one other function.
The fracture was defined as healed if trabeculations were
visible radiologically across the fracture line, with no sign
of AVN. Nonunion was defined as an absence of trabecula-
tion across the fracture line and/or redisplacement. In the
THR group, we analysed the radiographs for signs of loos-
ening of the components.19,20
Charnley’s numerical classification18 defines the clinical
state of the affected hip in regard to pain, movement and
ability to walk. Each feature is graded from 1 to 6, of which
1 is total disability and 6 the normal state. Although the
Charnley hip score is an ordinal variable, each value is pre-
sented as the mean.18 These results are shown in Table II.
The percentage of patients with the best scores (5 and 6) is
also shown.
THE JOURNAL OF BONE AND JOINT SURGERY
 INTERNAL FIXATION COMPARED WITH TOTAL HIP REPLACEMENT FOR DISPLACED FEMORAL NECK FRACTURES IN THE ELDERLY
The HRQoL was rated using the EQ-5D, which has five
categories: mobility, self-care, usual activities, pain or dis-
comfort and anxiety or depression.17 Each category is
divided into three degrees of severity: no problem, some
problems and major problems. Dolan et al21 used the Time
Trade-Off method to rate differing states of health in a large
UK population (UK EQ-5D Index Tariff). We used the pref-
erence scores (EQ-5D index scores) generated from this pop-
ulation when calculating the scores for our study
population. A value of 0 indicated death or a health state
worse than death, a value of 1 indicated full health. This is a
divergence from the UK EQ-5D Index Tariff in which some
health states were given a negative score, but the appropriate
scaling of negative scores is controversial.22,23 To validate
the method of rating the HRQoL before the injury and to
detect any recall bias, the EQ-5D index scores before the
fracture were compared with those of the age-matched
Swedish reference population. They were slightly better
(0.85) for the study age group of 70 to 79 years and for the
age group of 80 to 89 years (0.80) compared with the refer-
ence population (0.79 and 0.74, respectively).
In the outcome analysis, all patients remained in their
primary randomisation groups according to the intention-to-
treat principle, regardless of secondary procedures. There
was no significant difference between the THR and IF
groups regarding baseline data as shown in Table I.
Statistical analysis. The statistical software used was SPSS
(SPSS Inc, Chicago, Illinois) 11.0 for Windows. All scale
variables were tested for normality with the Kolmogorov-
Smirnov test. Student’s t-test was used for parametric scale
variables in independent groups and the Mann-Whitney U
test for non-parametric scale variables and ordinal variables
in independent groups. Nominal variables were tested by
the chi-squared test or Fisher’s exact test. All tests were
two-sided. The results were considered to be significant at p
< 0.05. Trend values (0.05 ≥ p < 0.1) are displayed; all other
values are reported as not significant (NS).
Results
Operative data. The mean operating time in the IF group,
including reduction, was 20 minutes (7 to 35). The operative
blood loss was 20 ml (0 to 100); three patients required
postoperative blood transfusions. The reduction was consid-
ered to be good in 46 of 53 patients (87%) and to be fair in
the remaining seven. The position of the screws was good in
51 of 53 patients (96%). There was no correlation between
the incidence of complications of fracture healing and the
accuracy of reduction or the position of the screws.
The mean operating time in the THR group was 102 min-
utes (40 to 152). The operative blood loss was 550 ml (150
to 1600) and 38 of 49 patients (77%) were given postopera-
tive blood transfusions (mean volume 640 ml (maximum
2400)). The position of the stem was considered to be good
in 48 of 49 patients (98%); in one patient the stem was in 8˚
of varus. The acetabular component was in a good position
VOL. 85-B, No. 3, APRIL 2003
383
1.0
0.9
Percentage surviving patients not re-operated
THR
0.8
0.7
0.6
0.5
0.4 IF
0.3
0.2
0.1
0.0
0 10 20 30
Time (months)
Fig. 2
Percentage of the 102 patients who are still alive and who have not been re-
operated upon with respect to time.
in 44 of 49 patients (90%). In five of the remaining patients
the lateral opening was increased by 5˚ to 8˚; one also had
6˚ of retroversion. The operated leg was lengthened by a
mean of 6 mm (-10 to 24).
General complications. Before the four-month follow-up,
five patients in the THR group developed general medical
complications: two superficial wound infections (negative
cultures; antibiotics for two weeks), two deep-venous
thromboses and one decubital ulcer. In the IF group, five
patients developed general medical complications: one pul-
monary embolus, one myocardial infarction and three
deaths. Of the total of 102 patients, 15 died during the
observation period, ten (19%) in the IF group and five
(10%) in the THR group (NS). The deceased patients were
older than the survivors, 83.8 ± 6.4 compared with 79.8 ±
5.7 years of age (p < 0.05) and had a lower HRQoL (EQ-5D
index score), 0.76 ± 0.13 compared with 0.84 ± 0.18 (p <
0.05). There were no significant differences in other base-
line data such as SPMSQ, gender, mobility, activities of
daily living (ADL) and comorbidity.
Surgical outcome. The surgical outcome is shown in
Figure 1. The number of surgical complications in the THR
group was two of 49 (4%); one patient with a malpositioned
acetabular component sustained three dislocations and was
eventually reoperated on six months after the primary sur-
gery. The acetabular component was repositioned and the
neck of the prosthesis was lengthened. No further disloca-
tion occurred. The patient had good function of the hip at
the 12-month follow-up but subsequently refused further
examination. He reported satisfactory hip function in a tele-
384 J. TIDERMARK, S. PONZER, O. SVENSSON, A. SÖDERQVIST, H. TÖRNKVIST

Table III. Deterioration in the mean (± SD) scores for quality of life (∆EQ-
5D index score) from before the fracture and follow-up for the 94 patients
followed for four months, the 92 for 12 months and the 84 for 24 months
Follow-up
(months) THR IF p value
0 to 4 -0.09 ± 0.24 -0.25 ± 0.25 <0.005
0 to 12 -0.09 ± 0.27 -0.23 ± 0.26 <0.05
0 to 24 -0.11 ± 0.28 -0.22 ± 0.24 <0.05

Table IV. Hip function according to the Charnley score for the 67 patients (42 in the THR group, 25 in
the IF group) without hip complications at final follow-up
THR IF
Percentage of Percentage of
patients with patients with
Follow-up Mean scores of 5 and 6 Mean scores of 5 and 6
(months) value (%) value (%) p value
Pain 4 5.7 88 5.0 63 <0.005
12 5.3 76 4.4 48 <0.01
24 5.6 83 5.1 60 0.062
Movement 4 5.0 65 4.7 55 NS
12 5.0 68 4.4 32 <0.005
24 4.9 68 4.6 50 NS
Walking 4 4.4 34 3.8 17 <0.05
12 4.7 51 4.0 28 <0.05
24 4.6 52 4.3 48 NS

phone interview 26 months after primary surgery. Another
patient fell six weeks after initial surgery and suffered a
periprosthetic fracture of the femoral shaft. This was fixed
internally and the final outcome was good. There were no
signs of radiological loosening of the components in any of
the patients at the final follow-up.
In the IF group, 19 of 53 (36%) patients sustained a com-
plication; 12 (23%) failed to unite and seven (13%) suffered
AVN. Second operations were needed in 22 of 53 patients
(42%). The procedures included replacement arthroplasty in
13 (four hemiarthroplasties and nine total arthroplasties), and
removal of screws in nine. Of those converted to arthroplasty,
three later died. The patients developing a complication did
not differ from those with uneventful healing regarding base-
line data (Table I). Local pain was the reason for removal of
screws from five out of 25 (20%) patients with uneventfully
healed fractures before the two-year follow-up. Two patients
failed to attend the 24-month follow-up after showing no
complication at 12 months. They were contacted by tele-
phone at 23 and 24 months, respectively; both reported con-
tinuing satisfactory function of the hip.
The complication rate, including reoperations, differed
between groups (p < 0.001). A life-table analysis of the sur-
viving patients who had not undergone secondary surgery, is
shown in Figure 2.
Functional outcome and HRQoL for all patients. There
were no differences in ADL between the groups at any of
the follow-ups. At four months, 89% in the IF group and
88% in the THR group were categorised as index A or B.
This compared with 91% and 85% at 12 months and 88%
and 90% at 24 months, respectively. Hip function was better
in the THR group (Table II).
Table V. Deterioration in the (mean ± SD) scores for quality of life (∆EQ-
5D index score) from before fracture and follow-up for the 67 patients (42 in
the THR group, 25 in the IF group) without hip complications at final
follow-up
Follow-up
(months) THR IF p value
0 to 4 -0.10 ± 0.25 -0.25 ± 0.22 <0.01
0 to 12 -0.09 ± 0.28 -0.21 ± 0.25 0.097
0 to 24 -0.10 ± 0.28 -0.17 ± 0.25 NS
The HRQoL, according to the EQ-5D index score, was
higher in the THR group at each follow-up, but the differ-
ences were only significant at four and 12 months (Fig. 3).
The difference in the EQ-5D index score, i.e. the change of
score between inclusion and each follow-up (4, 12 and 24
months), is shown in Table III. The decline in the EQ-5D
index score was more pronounced in the IF group.
Functional outcome and HRQoL for patients without hip
complications. At the final follow-up there remained 25 of
53 patients (47%) without a hip complication in the IF
group and 42 of 49 (86%) in the THR group (Fig. 1). There
were no differences in ADL between the groups at any of
the follow-up periods. At four months, 88% in the IF group
and 92% in the THR group were categorised as index A or
B, at 12 months 93% and 88%, and at 24 months 88% and
92%, respectively. The THR group had a significantly better
outcome regarding pain and walking ability at the four- and
12-month follow-up (Table IV). The HRQoL according to
EQ-5D did not differ significantly between groups (Fig. 4),
but the decrease in the EQ-5D index score was significantly
larger in the IF group between inclusion and the four-month
follow-up (Table V).
THE JOURNAL OF BONE AND JOINT SURGERY
 INTERNAL FIXATION COMPARED WITH TOTAL HIP REPLACEMENT FOR DISPLACED FEMORAL NECK FRACTURES IN THE ELDERLY 385

1.00 1.00
0.85 (0.13) 0.89 (0.12)
EQ-5D index score; mean (SD)

EQ-5D index score; mean (SD)

0.80 0.73 (0.20) 0.73 (0.25) 0.80 0.72 (0.21) 0.73 (0.24)
0.80 (0.22) 0.70 (0.28) 0.80 (0.23) 0.71 (0.27)
0.70 (0.28)
0.60 0.63 (0.27) 0.64 (0.28) 0.60 0.65 (0.19) 0.68 (0.24)
0.60 (0.22)

0.40 0.40
THR THR
IF IF
0.20 0.20

0.00 0.00
Before fracture, At 4 months, At 12 months, At 24 months, Before fracture, At 4 months, At 12 months, At 24 months,
ns p<0.005 p<0.05 ns ns p=0.071 ns ns
Time Time
Fig. 3 Fig. 4

The HRQoL for the 102 patients before fracture and the 94 available at the The HRQoL for the 67 patients (42 in the THR group, 25 in the IF group)
four-month, the 92 at the 12-month and the 84 at the 24-month follow-up. with no complications at final follow-up.

Discussion Comparing THR with IF for a displaced fracture of the

femoral neck in the elderly, yielded a significantly lower
The treatment of displaced fractures of the femoral neck is complication rate for THR, 4% compared with 36%, and an
controversial, especially from an international perspective. infrequent need for secondary surgery, 4% compared with
For many years, the standard treatment in Sweden, in spite 42%. In addition, THR produced significantly better hip
of high rates of complications, has been IF, based mainly on function and a better health-related quality of life. When
the argument that retaining the patient’s femoral head can considering only patients with an uneventful postoperative
always give better function of the hip than an arthroplasty. course, the hip function after a THR and the HRQoL in the
In most other European countries, the preferred treatment early postoperative phase, at four months, were better.
has been a primary arthroplasty, usually with a unipolar or In our study, the results after IF revealed a complication
bipolar femoral prosthesis. Most studies on this subject rate of 36%. This is equal to or better than those in most
focus on complications and the need for further surgery other studies,3 including two recent randomised controlled
rather than function of the hip, and almost none gave an trials which compared IF with THR in which the rate of
assessment of the quality of life. complications at the two-year follow-up was almost 50%.8,9
The population of elderly patients with fractures of the The fracture reduction and screw position, according to the
femoral neck is heterogeneous and contains a spectrum of current best practice, were optimal in most patients and yet
patients ranging from the independent, healthy subject with the complication rate was unacceptably high. Rehnberg and
high functional demands, to the institutionalised, cogni- Olerud,24 using the same implant (Olmed screw), reported
tively impaired and bedridden patient. The techniques cur- complications of fracture healing (all AVN) in only 12%, in
rently available, including IF, unipolar arthroplasty, bipolar a study of 44 consecutive patients 43 of whom had a dis-
arthroplasty and total hip arthroplasty, have different out- placed fracture. These remarkably good results from one
comes and differing risk profiles. The good long-term particular surgeon could not be reproduced in a randomised
results of THR are confirmed in a recent study10 at 13 years controlled trial from the same institution, in which the com-
of the same population as in the study by Skinner et al6 plication rate at one year was 21% using the same
comparing IF, hemiarthroplasty and THR, in which the revi- implant.25 By comparison, in our study the rate was 23% at
sion rate was 33%, 24% and 7%, respectively. Hip function, one year. We believe that the results after IF, in osteoporotic
according to the Harris hip score, was best in the THR elderly patients with a displaced fracture of the femoral
group and worst in the hemiarthroplasty group. neck, cannot be significantly improved with current fixation
We believe that the displaced fracture of the femoral techniques.
neck merits a more patient-related rather than diagnosis- Even after a successfully healed displaced fracture,
related approach. In order to achieve this, we need ran- removal of the screw is often required in 20% of our
domised, controlled clinical trials in which the outcome patients. This issue is rarely addressed in most studies of IF,
analysis also includes the patient’s perspective of his or her but Sernbo et al26 reported removal of the implant after
quality of life. Our aim was to analyse the outcome after a healing in approximately 7% of patients. In the ten-year
displaced fracture of the femoral neck in an elderly, rela- follow-up by Jonsson et al27 the implant was removed from
tively healthy, independent patient, randomly selected to 28% of the patients. The number of patients having implant-
have a THR or IF. related problems is probably even higher. A treatment with a
VOL. 85-B, No. 3, APRIL 2003
386 J. TIDERMARK, S. PONZER, O. SVENSSON, A. SÖDERQVIST, H. TÖRNKVIST
built-in need for secondary surgery, albeit minor proce-
dures, is not ideal.
Most complications of fracture healing require a second-
ary arthroplasty, provided that the follow-up is continuous.
Without regular follow-up there is a risk that elderly patients
will not return for help, instead adapting to reduced function
and quality of life. Arthroplasty after failed IF was carried
out in 68% of the patients (for nonunion in 92%; for AVN in
29%). The indications for an arthroplasty are nearly always
relative, aimed at improving function and quality of life, and
must be balanced against surgical risks. Some of the
patients were reoperated upon by surgeons who were not
involved in the study, which explains the variety of proce-
dures: THR in 69% and hemiarthroplasty in 31%. Of the
patients with a complication of fracture healing, 32% were
not reoperated on during the period of the study.
The rate of dislocation of 2% after primary THR in our
study compares favourably with those in previous
reports.3,6,8,9 In our experience, the anterolateral approach
is the best approach for optimising stability, which is the
primary goal in patients with fracture of the femoral
neck.28,29 Recent randomised, controlled trials of THR for
fracture of the femoral neck used the posterolateral
approach and reported a dislocation rate of 14% and 22%.8,9
Another factor in avoiding dislocations was the exclusion of
patients with severe cognitive dysfunction. In the series
from Johansson et al,8 the dislocation rate in patients with
mental dysfunction was 32% compared with 12% in men-
tally competent patients. Finally, we believe that the experi-
ence of the surgeons and the design of the prosthesis
improved the outcome. Deep infection, previously a signifi-
cant problem, now seems to be rare with modern techniques
and antibiotic prophylaxis.7-9
The rate of general complications was similar in both
groups. The two-year mortality rate (15%) is comparable
with or lower than that in most other studies of internal fixa-
tion and lower than in recent randomised, controlled trials
which compare IF with THR. Johansson et al8 reported a
two-year rate of 33%, probably due to the selection of
slightly older (mean age 84 years) and cognitively impaired
patients (mental dysfunction in 45%). Neander9 reported a
mortality rate at two years of 24% and, as in our study, there
was a tendency to increased mortality in the IF group (28%)
compared with the THR group (19%).
The surgical procedure itself does not increase the mor-
tality rate. No patient in the THR group died during the first
four months after surgery. One reason for the increased
mortality rate in the IF group could be a selection bias, but
that seems less likely since the randomisation groups in our
study were comparable regarding baseline data. A more
probable explanation is that the functional deterioration and
pain impair the patient’s general condition and thus increase
mortality. Those who died were four years older and had a
lower HRQoL than survivors.
The number of patients lost to follow-up was low (3%).
All attended the follow-up at 12 months and were inter-
viewed on the telephone at the time of the 24-month follow-
up. Their outcome is reported and the fact that they did not
attend the latter should not affect the interpretation of the
results.
We studied a rather healthy group of patients for several
reasons. Only relatively healthy elderly patients with high
functional demands and a relatively long life expectancy can
fully enjoy the advantages of a THR. Secondly, patients
with cognitive dysfunction have a high rate of disloca-
tion.3,8 We excluded only patients with severe cognitive
impairment (SPMSQ <3), a group constituting about 10%
of all hip fracture patients admitted from independent living
circumstances.30
The age level is arbitrary. We chose ≥70 years and, in
previous randomised, controlled trials, ≥65 years and ≥75
years.8,9 The reason for not undertaking THR in younger
patients is the cumulative risk of revision because of aseptic
loosening, but this risk seems to be minor from a ten-year
perspective. The percentage of Swedish patients with hip
fractures not revised 11 years after a THR is 98%.31 The
expected mean survival of a 70-year-old Swedish woman is
16 years and of a man 13 years, and although the mortality
is higher after a hip fracture, patients surviving the first year
have an expected mean survival comparable with that of the
rest of the population.32,33 In addition, the results after IF
are better in younger patients.34 The optimal lower age for a
THR is yet to be determined, but current results support the
choice of 70 years. We did not set an upper age limit, but the
other inclusion criteria excluded the oldest and most fragile
patients, that is to say the approach considers biological
rather than chronological age.
Most scores evaluating hip function have been validated
for patients with THR after degenerative joint disease, but
they have also recently been presented in studies comparing
THR and IF in patients with fractures of the femoral
neck.8,9,18,35 We used the Charnley hip score because it is
well validated and the outcome is indicated in three impor-
tant criteria regarding hip function. The values for pain and
movement were relatively stable between all follow-up
occasions, but were on a higher level for the THR group.
The values for walking improved between the four- and 12-
month follow-up periods and stabilised thereafter on a
higher level for the THR group and for the IF patients with
healed fractures. The explanation of the generally lower
values in walking compared with those in studies on
patients with THR after osteoarthritis is probably related to
the older age and more frequent comorbidities in the hip
fracture population.18 Even before the fracture, 22% of the
patients used some sort of walking aid for other reasons
than hip disorders.
A quality-of-life assessment provides a general outcome
from the patient’s point of view and provides information
about the effect of the injury. The EuroQol is brief and easy
to use even in elderly patients.4,36,40 It can also be used for
calculating quality-adjusted life-years (QUALY), as
required for health-care evaluations. The patient’s ability to
THE JOURNAL OF BONE AND JOINT SURGERY
 INTERNAL FIXATION COMPARED WITH TOTAL HIP REPLACEMENT FOR DISPLACED FEMORAL NECK FRACTURES IN THE ELDERLY
recall their health status before the hip fracture may be
questioned40 but the rated prefracture EQ-5D index score of
our patients showed correlation with an age-matched
Swedish reference population.23 The EQ-5D index score
before injury was slightly higher than the reference popula-
tion, indicating that the study population was relatively
healthy.
The higher complication and mortality rates in the IF
group were associated with a higher HRQoL in those with
an uneventful postoperative course, which may affect com-
parisons between the groups. We also investigated the alter-
ation in the quality of life by comparing the prefracture and
follow-up values. The decrease in the quality of life (decline
in EQ-5D index score) over the two-year period for patients
with an uneventfully healed (IF) displaced fracture, was
slightly lower than previously reported by us, 0.17 com-
pared with 0.24, but the inclusion criteria for these studies
were slightly different.5 The decrease in the EQ-5D index
score after THR was comparable with the decrease after an
uneventfully healed undisplaced fracture, -0.10 in both stud-
ies.5
The limited number of surgeons involved in our study is
both a weakness and a strength. In regard to the latter both
procedures were carried out by experienced surgeons,
thereby providing a fair comparison of the methods and the
weakness is in terms of the ability to make generalisations.
We believe our findings are amenable to generalisation
because, in recent randomised controlled trials in which
procedures were carried out by unselected surgeons, the
conclusions were still that THR is the method of choice for
treatment, with a lower complication rate and a better func-
tional outcome.8,9
In conclusion, the results of our study strongly support
the view that a primary THR is preferable for an elderly, rel-
atively healthy, lucid patient with a displaced fracture of the
femoral neck. We found a lower complication rate than after
IF, and the outcome regarding hip function and HRQoL is
generally better. On the other hand, for the group of patients
with a successfully healed fracture, two years after IF, hip
function and HRQoL are comparable with those achieved
after a primary THR. IF can be justified only if the follow-
up is scrupulous and scheduled at regular intervals with
urgent conversion to an arthroplasty if fracture healing fails
for any reason.
This study was supported in part by grants from the Trugg-Hansa Insurance
Company, Swedish Society for Medical Research, the Swedish Orthopaedic
Association and the Stockholm County Council.
Although none of the authors have received or will receive benefits for
personal or professional use from a commercial party related directly or in-
directly to the subject of this article, benefits have been or will be received
but are directed solely to a research fund, foundation, educational institu-
tion, or other non-profit institution with which one or more of the authors is
associated.
References
1. Cooper C, Campion G, Melton J III. Hip fractures in the elderly: a
world-wide projection. Osteoporos Int 1992;2:285-9.
2. Garden RS. Low-angle fixation in fractures of the femoral neck. J Bone
Joint Surg [Br] 1961;43-B:647-63.
VOL. 85-B, No. 3, APRIL 2003
387
3. Lu-Yao GL, Keller RB, Littenberg B, Wennberg JE. Outcomes
after displaced fractures of the femoral neck: a meta-analysis of one
hundred and six published reports. J Bone Joint Surg [Am] 1994;76-
A:15-25.
4. Tidermark J, Zethraeus N, Svensson O, Törnkvist H, Ponzer S.
Femoral neck fractures in the elderly: functional outcome and quality of
life according to the EuroQol. Qual Life Res 2002;11:473-81.
5. Tidermark J, Zethraeus N, Svensson O, Törnkvist H, Ponzer S.
Quality of life related to fracture displacement among elderly patients
with femoral neck fractures treated with internal fixation. J Orthop
Trauma 2002;16:34-8.
6. Skinner P, Riley D, Ellery J, et al. Displaced subcapital fractures of
the femur: a prospective randomised comparison of internal fixation,
hemiarthroplasty and total hip replacement. Injury 1989;20:291-3.
7. Jonsson B, Sernbo I, Carlsson A, Fredin H, Johnell O. Social func-
tion after cervical hip fracture: a comparison of hook-pins and total hip
replacement in 47 patients. Acta Orthop Scand 1996;67:431-4.
8. Johansson T, Jacobsson SA, Ivarsson I, Knutsson A, Wahlström O.
Internal fixation versus total hip arthroplasty in the treatment of dis-
placed femoral neck fractures: a prospective randomised study of 100
hips. Acta Orthop Scand 2000;71:597-602.
9. Neander G. Displaced femoral neck fractures: studies on osteosynthe-
sis and total hip arthroplasty. Thesis 2000; Karolinska Institutet. http://
diss.hib.ki.sc/2000/91-628-4167-X/
10. Ravikumar KJ, Marsh G. Internal fixation versus hemiarthroplasty
versus total hip arthroplasty for displaced subcapital fractures of femur:
13-year results of a prospective randomised study. Injury 2000;31:
793-7.
11. Rogmark C, Carlsson A, Johnell O, Sernbo I. A prospective ran-
domised trial of internal fixation versus arthroplasty for displaced frac-
tures of the neck of the femur: functional outcome for 450 patients at
two years. J Bone Joint Surg [Br] 2002;84-B:183-8.
12. Pfeiffer E. A short portable mental status questionnaire for the assess-
ment of organic brain deficit in elderly patients. J Am Geriatr Soc
1975;23:433-41.
13. Lindequist S, Törnkvist H. Quality of reduction and cortical screw
support in femoral neck fractures: an analysis of 72 fractures with a new
computerized measuring method. J Orthop Trauma 1995;9:215-21.
14. Hardinge K. The direct lateral approach to the hip. J Bone Joint Surg
[Br] 1982;64-B:17-9.
15. Ceder L, Thorngren KG, Wallden B. Prognostic indicators and early
home rehabilitation in elderly patients with hip fractures. Clin Orthop
1980;152:173-84.
16. Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW. Studies of
illness in the aged: the index of ADL: a standardized measure of biolog-
ical and psychological function. JAMA 1963;185:94-9.
17. Brooks R. EuroQol: the current state of play. Health Policy 1996;37:53-
72.
18. Charnley J. The long-term results of low-friction arthroplasty of the hip
performed as a primary intervention. J Bone Joint Surg [Br] 1972;54-
B:61-76.
19. Gruen TA, McNeice GM, Amstutz HC. “Modes of failure” of cement-
ed stem-type femoral components: a radiographic analysis of loosening.
Clin Orthop 1979;141:17-27.
20. DeLee JG, Charnley J. Radiological demarcation of cemented sockets
in total hip replacement. Clin Orthop 1976;121:20-32.
21. Dolan P, Gudex C, Kind P, Williams A. The time trade-off method:
results from a general population study. Health Econ 1996;5:141-54.
22. Macran S, Kind P. “Death” and the valuation of health-related quality
of life. Med Care 2001;39:217-27.
23. Burström K, Johannesson M, Diderichsen F. Swedish population
health-related quality of life results using the EQ-5D. Qual Life Res
2001;10:621-35.
24. Rehnberg L, Olerud C. Subchondral screw fixation for femoral neck
fractures. J Bone Joint Surg [Br] 1989;71-B:178-80.
25. Rehnberg L, Olerud C. Fixation of femoral neck fractures: comparison
of the Uppsala and von Bahr screws. Acta Orthop Scand, 1989;60:579-
84.
26. Sernbo I, Johnell O, Bååth L, Nilsson JA. Internal fixation of 410 cer-
vical hip fractures: a randomised comparison of a single nail versus two
hook-pins. Acta Orthop Scand 1990;61:411-4.
27. Jonsson B, Johnell O, Redlund-Johnell I, Sernbo I. Function 10 years
after hip fracture: 74 patients after internal fixation. Acta Orthop Scand
1993;64:645-6.
388 J. TIDERMARK, S. PONZER, O. SVENSSON, A. SÖDERQVIST, H. TÖRNKVIST
28. Woo RYG, Morrey BF. Dislocations after total hip arthroplasty. J Bone
Joint Surg [Am] 1982;64-A:1295-306.
29. Taine WH, Armour PC. Primary total hip replacement for displaced
subcapital fractures of the femur. J Bone Joint Surg [Br] 1985;67-
B:214-7.
30. Strömberg L, Lindgren U, Nordin C, Öhlen G, Svensson O. The ap-
pearance and disappearance of cognitive impairment in elderly patients
during treatment for hip fracture. Scand J Caring Sci 1997;11:167-75.
31. The Swedish National Hip Arthroplasty Registry. 2002. http://
www.jru.orthop.gu.se/
32. Statistics Sweden. 2002. http://www.scb.se
33. Clayer MT, Bauze RJ. Morbidity and mortality following fractures of
the femoral neck and trochanteric region: analysis of risk factors. J Trau-
ma 1989;29:1673-8.
34. Broos PL, Vercruysse R, Fourneau I, Driesen R, Stappaerts KH.
Unstable femoral neck fractures in young adults: treatment with the AO
130-degree blade plate. J Orthop Trauma 1998;12:235-9.
35. Harris WH. Traumatic arthritis of the hip after dislocation and acetab-
ular fracture: treatment by mold arthroplasty. J Bone Joint Surg [Am]
1969;51-A:737-55.
36. Brazier JE, Walters SJ, Nicholl JP, Kohler B. Using the SF-36 and
EuroQol on an elderly population. Qual Life Res 1996;5:195-204.
37. Hurst NP, Kind P, Ruta D, Hunter M, Stubbings A. Measuring
health-related quality of life in rheumatoid arthritis: validity, responsive-
ness and reliability of EuroQol (EQ-5D). Br J Rheumatol 1997;36:551-
9.
38. Dorman P, Slattery J, Farrell B, Dennis M, Sandercock P. Qualitative
comparison of the reliability of health status assessments with the Euro-
Qol and SF-36 questionnaires after stroke. United Kingdom Collabora-
tors in the International Stroke Trial. Stroke 1998;29:63-8.
39. Harper R, Brazier JE, Waterhouse JC, et al. Comparison of outcome
measures for patients with chronic obstructive pulmonary disease
(COPD) in an outpatients setting. Thorax 1997;52:879-87.
40. Williams A. The role of the EuroQol Instrument in QUALY calcula-
tions. York: The University of York, Centre for Health Economics, 1995.
THE JOURNAL OF BONE AND JOINT SURGERY