ACUTE HAEMARTHROSIS OF THE KNEE IN ATHLETES

A PROSPECTIVE STUDY OF 106 CASES

NICOLA MAFFULLI, PETER M. BINFIELD, JOHN B. KING, CHRISTOPHER J. GOOD

From Newham General Hospital, London, England

We made a prospective arthroscopic study of 106 skeletally The most common mechanism producing a traumatic
mature male sportsmen with an average age of 28.35 years haemarthrosis of the knee is a twist of the flexed joint,
(16.8 to 44) who presented with an acute haemarthrosis of although many patients are uncertain of the position or
the knee due to sporting activities. We excluded those with the direction of movement at the time of injury (Noyes et
patellar dislocations, radiographic bone injuries, extra- al 1980). Patients may be unable to continue playing their
articular ligamentous lesions or a previous injury to the sport (King and Aitken 1988), but about two-thirds
same joint. describe merely a popping sensation and swelling of the
The anterior cruciate ligament (ACL) was intact in 35 joint within two hours (Noyes et al 1980, 1983).
patients, partially disrupted in 28 and completely ruptured Such a story alone may indicate the diagnosis
in 43. In the patients with an ACL lesion, associated injuries (Feagin, Abbott and Rokous 1972), and early clinical
included meniscal tears (17 patients), cartilaginous loose examination is unreliable (Noyes et al 1980; Rand 1984;
bodies (6), and minimal osteochondral fractures of the Simonsen et al 1984). A tense effusion with muscle
patella (2), the tibial plateau (3) or the femoral condyle (9). guarding severely limits the range ofmotion, and stability
We found no age-related trend in the pattern of ACL is therefore difficult to evaluate (DeHaven 1980). For
injuries. Isolated injuries included one small osteochondral these reasons only 9% to 29% ofanterior cruciate ligament
fracture of the patella, and one partial and one total (ACL) tears which are found arthroscopically have been
disruption ofthe posterior cruciate ligament. Three patients previously diagnosed clinically (DeHaven 1980 ; Noyes
had cartilaginous loose bodies, and no injury was detected et al 1980).
in five. Arthroscopy provides reliable information and is
Acute traumatic haemarthrosis indicates a serious indicated less by what it can achieve than by the failure
ligament injury until proved otherwise, and arthroscopy is of other diagnostic measures (Glinz, Segantini and K#{228}gi
needed to complement careful history and clinical exami- 1980). Previous reports on traumatic haemarthrosis from
nation. All cases with a tense effusion developing within 12 North America (DeHaven 1980; Noyes et al 1980) and
hours of injury should have an aspiration. If haemarthrosis continental Europe are available (Gillquist, Hagberg and
is confirmed, urgent admission and arthroscopy are Oretorp 1977). From the UK, Jam, Swanson and
indicated. Murdoch (1983), reporting on a mixed sedentary and
athletic population, found that only 17% had an ACL
J Bone Joint Surg [Br] 1993 ; 75-B :945-9.
tear, but they did not perform routine arthroscopy and
Received 5 February 1993 ; Accepted 22 April 1993
their main line oftreatment was aspiration and splintage,
which is no longer an accepted practice (Mariani, Puddu
and Ferretti 1982).
Our aim was to make a longitudinal study of athletes
with acute haemarthrosis, using arthroscopy and exami-
nation under anaesthesia (EUA). Our hypothesis was
that an acute traumatic haemarthrosis in an athlete
implies significant intra-articular injury which should be
accurately diagnosed.
N. Maffulli, MD, PhD, Senior Registrar and Clinical Lecturer in
Orthopaedics
P. M. Binfield, FRCS, Career Registrar in Orthopaedics
J. B. King, FRCS, Senior Lecturer and Consultant in Orthopaedics PATIENTS AND METHODS
C. J. Good, FRCS, Consultant in Orthopaedics
Department of Orthopaedics, Newham General Hospital, Glen Road, From January 1986 to April 1992, 106 male, skeletally
Plaistow, London E13 85L, UK.
mature athletes with acute haemarthrosis sustained
Correspondence should be sent to Dr N. Maffulli at the Department of
Orthopaedic Surgery, UniversityofAberdeen MedicalSchool, Polwarth during a sporting activity were admitted to Newham
Building, Foresterhill, Aberdeen AB9 2ZD, UK. General Hospital, London. Patients with a history of a
©l993 British Editorial Society ofBone and Joint Surgery previous injury to the same joint were excluded, as were
0301 -620X/93/6656 $2.00
those with obvious patellar dislocation, a radiologically

VOL. 75-B, No. 6, NOVEMBER 1993 945

946 N. MAFFULLI, P. M. BINFIELD, J. B. KING, C. J. GOOD

evident bone injury, or an extra-articular ligamentous diagnosis of a meniscal injury was made. Twenty-two
lesion. The average age of the patients was 28.35 ± SD patients were referred immediately to an orthopaedic
8.13 (16.8 to 44, Fig. I). The right knee was injured in 81 surgeon without an initial aspiration.
ofthe 106 patients. The other 1 3 patients were referred to a trauma
clinic without joint aspiration and without diagnosis.
The orthopaedic management included a full clinical
examination, and aspiration if this had not already been
performed. Patients were then offered admission with a
view to arthroscopy within the next 24 hours.
Examination under anaesthesia. No patient showed
isolated medial or isolated lateral laxity on examination
with or without anaesthesia. The Lachman test (Torg,
Conrad and Kalen 1976) and the pivot shift test (Galway,
Beaupr#{233}and MacIntosh 1972) were performed under full
general anaesthesia before a tourniquet was applied.
Arthroscopy. All EUAs and arthroscopies were per-
formed or supervised by one of the senior authors (JBK
and CJG). At arthroscopy under tourniquet, antero-
..
medial, anterolateral, posteromedial and posterolateral
Age at presentation (years)
entry portals were used as required through vertical (83)
Fig. 1
or horizontal 5 mm stab wounds. Tourniquets were not
Age distribution of 106 athletes with haemarthrosis ofthe knee. used in five patients with sickle-cell trait, as determined
by family history and Hb electrophoresis in some ethnic
groups. In 81 patients gravity irrigation with normal
The sports played at the time of injury were soccer saline and a medial suprapatellar drain were used ; the
(33), rugby (15), hockey (14), American football (12), others had gravity irrigation without a drain. Normal
track-and-field athletics (12), combat sports (10), gym- saline was used to wash out the joint.
nastics (7),and cricket (6). Sports were played at The wounds were closed with a single Dexon or
recreational level (47), county level (12), national level Ethilon suture in 34 patients and unsutured in the
(28), and international level ( 19). This distribution reflects remainder. A modified Robert Jones compression band-
the association oftwo ofthe authors (NM and JBK) with age was applied before removing the tourniquet, and
a National Sports Centre and various national sports patients were discharged on the same day when they
organisations. were able to perform an unaided straight-leg raise. When
Mechanism of injury. In 73 patients the injury had been a drain had been left in place, it was removed in the day-
caused by a non-contact event which involved twisting, care unit before discharge. Outpatient reviews were at
turning or jumping off the planted foot. In 18 patients two and at six weeks, or for any clinical problem.
there had been a direct blow on the upper third of the ACL tears. A blunt hook-shaped probe was used to
lower leg, with the knee slightly flexed. The other 15 examine the ACL through an appropriate entry portal.
athletes were unable to describe the position oftheir knee Any synovial tissue which obscured the cruciate fibres
at the time of injury. was lifted or partially removed (Noyes et al 1980).
Symptoms. The most common symptom at the time of Tension in the fibres of the ACL was judged by pulling
injury was sudden pain with a popping sensation and with the hook. ACL tears were classified as partial or
collapse of the leg. This was described by 72 patients, complete on the basis of visual inspection and probing.
and was followed by a fall to the ground in 26 of them. Meniscal Meniscal
injuries. tears have been classified by
Sixty-four patients were able to resume the sporting appearance as horizontal, longitudinal, vertical, etc
activity immediately after injury, but had to stop later. (Barber 1992), by the thickness of the tear (full or partial
Swelling and joint effusion were evident within two split) and by location in the meniscus (peripheral, central,
hours in 83 patients, and within 1 2 hours in all of them. etc: Irvine and Glasgow 1992). We classified meniscal
When first seen, additional complaints included giving tears according to the involved side and the shape and
way (24), an insecure feeling (2 1 ) and true locking of the the location within the meniscus.
knee, defined as inability to extend it fully (16). All data were collected prospectively on a proforma
Before arthroscopy. The patients were first seen in the (Fig. 2). This was completed in triplicate by the operating
Accident and Emergency Department at an average of surgeon at the time of the procedure, providing one copy
3.2 ± 1 .6 days (0 to 4) from the injury. Anteroposterior for the patient’s general practitioner, one for the operating
and lateral radiographs were taken in all cases. All had notes, and one for this study.
an effusion, and in 71 aspiration by a casualty officer The data were entered into a dBIII data base on an
confirmed the presence of a haemarthrosis. In 28 the IBM-compatible computer and were analysed using

THE JOURNAL OF BONE AND JOINT SURGERY

 ACUTE HAEMARTHROSIS OF THE KNEE IN ATHLETES 947

WHITE - Records
KNEE ARTHROSCOPY RECORD PINK - J.B.K
Dear Dr. [G P 1 YELLOW - G.P.

I enclose a copy of your patient’s arthroscopy record. Yours sincerely.

Name No Date

Address: Phone

Surgeon Age Side Hospital

COMPLAINTS: Trauma I I Overuse Spontaneous SPORT I

Recreatjon

PAIN GIVING LOCKING ACCIDENT

ANT SPONTANEOUS CANNOTEXT CONTACT
IMMEDIATE SWELLING
POST TWISTING FLEX - DISABILITY
MED STAIRS FIXED
-. SWELLING
LAT
PARAPAT M
RUNNING DELAYED
-.-.DISABILITY
L SWELLING
NON CONTACT
STAIRS SPORTSNOW ALWAYS
SITTING I INTERMITTENT- IMMEDIATESWELLING
DISABILITY
SQUATTING HIGH CLASS AFTER EPISODE
TWISTING RECREATION SWELLING
DELAYED
AT REST DIFFICULT DISARIUTY
ATNIGHT IMPOSSIBLE WORK(
OK DIFF IMPOSSIBLE

EXAMINATION
EXAMINATION ASLEEP 0-3 (RECOGNISES v/ OK X. NOTOK o NOTSEEN
FLUID ANT DRAW FLUID ANT DRAW HYPER EXIT( ) I NORMAL
VALGUSO NEUTRAL VALGUSO NEUTRAL PATELLATRACK ( HIGH
VARUSO ER VARUSO ER I LOW
VALGUS 30 IR VALGUS 30 IR
VARUS 30 POST DRAW VARUS 30 POST DRAW PAIN
LACHMAN POST LAT DRAW LACHMAN POST LAT DRAW FROGEVE
PIVOT REVERSE PIVOT PIVOT REVERSE PIVOT APPREHENSION
OTHER LOCKED LATTHRUST LOCKED CLARKE
REC/EXTROT PATELLADISLOC TENDERI KNOCKKNEE FXAMINFP

BAYONET CYST-MEN BOWLEG

REC/EXTROT SM STRAIGHT
BAKER

,scoPY SCOPE I I DRAIN V/N FLUID - CLEAR, SEROUS, BLOOD PORTALS

NOT OK DRAW ON CHART - DOTTED LINES

LM
cPP

LAT RECESS
MED RECESS
ACL
PCL
PLICA _ -

FATPAD _ -

SYNOVIUM _ -

PATELLAMED
LAT
FACET _

TIBIAMED _ -
GRADE 0-3
LAT
FEMUR MED _

LAT

LOOSE BODY NO WHERE

OCD WHERE

PROCEDURE DRAW
#{149} ON CHART - FULL LINES
0 P D DATE

DESCRIBE

Fig. 2

The proforma used to record details ofeach patient.

VOL. 75-B, No. 6, NOVEMBER 1993

948 N. MAFFULLI, P. M. BINFIELD, J. B. KING, C. J. GOOD

Systat (Leland 1988) for chi-squared tests and chi-square Table I. Results of examination under anaes-
thesia of 7 1 sportsmen with haemarthrosis and
for trend tests. Significance was set at the 0.05 level. ACL injury

Partial tear Complete tear

RESULTS (n=28) (n=43)

Lachman test
Examination under anaesthesia. In knees with complete +ve 21 34
disruption of the ACL, both Lachman and pivot shift -ye 7 9
tests were reliable (Table I). In those with only partial
Pivot shift test
rupture of the ACL, the pivot shift test was more reliable +ve 9 43
-ye 19
than the Lachman test. There were no false-positive
results.
ACL lesions (Table II). The ACL was intact in 35 knees, Table II. ACL lesions and associated injuries in
106 sportsmen with haemarthrosis
partially ruptured in 28 and completely disrupted in 43.
In 29 patients, the partial or complete disruption of the Lesion/injury Number
ACL was an isolated injury. In the other 42 patients with Intact ACL 35
an ACL injury, associated injuries included meniscal tear
Partial disruption of ACL 28
(1 7), cartilaginous loose bodies (6), or minimal osteochon- Isolated partial disruption of ACL 12
dral fracture of the patella (2), the tibial plateau (3), or
Complete disruption of ACL 43
the femoral condyles (9). We found no age-related trend Isolated complete disruption of ACL 17
in the pattern of ACL-associated injuries, or in the
Meniscal tears I7
development of a partial or complete ACL lesion (chi-
square for trend test). Cartilaginous loose bodies 6

Meniscal lesions. Of the 16 patients with a clinically Osteochondral fractures

locked knee, I 2 had a meniscal tear and four had an Patella 2
Tibial plateau 3
isolated ACL rupture. All 1 7 meniscal injuries ( 1 1 lateral, Femoral condyle 9
5 medial, one of both menisci) were associated with
Partial disruption of PCLt I
either total or partial ACL disruption. The lateral
meniscus was injured significantly more often (p = Complete disruption of PCL I

0.042, chi-squared test). There were nine bucket-handle S anterior cruciate ligament
tears (six lateral); all the others were posterior-horn tears. t posterior cruciate ligament
In six cases, an isolated peripheral tear was the cause of
the haemarthrosis (DeHaven 1980). These were repaired documented (Gillquist et al 1977; Lysholm et al 1981)
arthroscopically. but in some reports of ACL injury only 20% were caused
Three patients required arthrotomy after arthros- by sporting activities and 60% by simple falls (Jam et al
copy, two for subtotal lateral meniscectomy, and one for 1983). Road-traffic accidents are a less common cause of
partial medial meniscectomy. The other eight patients knee injuries (Lysholm et al 1981).
with meniscal tears had arthroscopic partial meniscecto- In all our patients, the haemarthrosis developed
mies. within 12 hours of an injury. It is usually possible to
Haemarthrosis with no ACL lesion. No intra-articular differentiate these by the history from minor, non-
injury could be detected in five patients, and it was haemorrhagic effusions that occur 24 hours or later after
assumed that the haemarthrosis was due to synovial injury.
tears. Apart from the six isolated peripheral meniscal A significant proportion of our patients, 64 of 106,
tears, other isolated injuries which appeared to have with ACL disruption were able to resume sporting
caused haemarthrosis were one small osteochondral activity immediately after their injury: some patients
fracture of the patella, and one partial and one total had little or no initial pain. The serious nature of the
disruption of the posterior cruciate ligament. Cartila- injury may not be apparent, and this is particularly true
ginous loose bodies were found in three patients and of non-contact twisting injuries.
removed arthroscopically. Examination of an acutely swollen knee is difficult
and unrewarding, but the easiest and most reliable test is
said to be that of Lachman (Torg et al 1976). The result
DISCUSSION
of this test may be subjective in that it relates to the
The average age (28.35 ± 8. 13)in ourseries is comparable hardness of the stop (King and Aitken 1988). Our results
with that of other studies, but our upper age (44 years) is show that a positive Lachman test with a negative pivot
lower than that in most other reports (O’Connor 1974; shift test are indicative of partial ACL rupture. The two
Gillquist et al 1977; Lysholm, Gillquist and Liljedahl tests do not give significantly different results in complete
1981 ; Jam et al 1983; Harilainen et al 1988). The high ACL tears, but we found the pivot shift test to be slightly
incidence of knee injury during sporting activities is well more sensitive.

THE JOURNAL OF BONE AND JOINT SURGERY

 ACUTE HAEMARTHROSIS OF THE KNEE IN ATHLETES 949

Acute traumatic haemarthrosis is an indication for et al 1981 ; Harilainen et al 1988). The proportion with a
urgent arthroscopy with the aim of establishing a partial ACL tear (28 of 7 1 , 39%) is also similar to that of
complete and precise diagnosis (O’Connor 1974; Noyes other series, 28% of 84 patients with haemarthrosis
et al 1980; Rand 1984; Harilainen et al 1988). Delay in (Noyes et al 1980) and 34% in a series of ACL ruptures
arthroscopy may allow synovial hypertrophy, which (Liljedahl, Lindvall and Wetterfors 1965). The presenta-
makes it difficult to visualise the ACL. Avulsion of the tion of ACL disruptions as medial meniscal injuries (four
ACL from its femoral attachment is visible for up to two of our patients) has also been reported (Jackson and
weeks, but by then mid-ligament tears are difficult to see Dandy 1976; Farquharson-Roberts and Osborne 1983).
because synovium conceals the torn ends (Noyes et al We conclude that acute traumatic haemarthrosis
1980). Urgent arthroscopy for acute traumatic haem- should be regarded as due to a serious ligament injury
arthrosis is therefore justified. Even if immediate surgical until proved otherwise. Arthroscopy is needed to comple-
repair is not needed, full assessment is important in ment a careful history and clinical examination. Patients
planning management (King and Aitken 1988). Partial with a tense effusion of the knee within 1 2 hours of an
rupture of the ACL may lead to a total rupture and acute injury should have an aspiration. If this confirms a
instability, and therefore these patients require careful haemarthrosis the patient should be referred urgently
follow-up (King 1991). with a view to admission and arthroscopy.
We found partial or complete rupture of the ACL in We thank Mr 0. A. Jamall, FRCS Ed, Mr J. K. Klosok, FRCS G, and
70% of our series, agreeing with other studies of athletes Mr B. Levack, FRCS, for allowing us to study patients under their care.
No benefits in any form have been received or will be received
(DeHaven 1980; Noyes et al 1980, 1989; King 1991). from a commercial party related directly or indirectly to the subject of
Series of mixed origin show a lower incidence (Lysholm this article.

REFERENCES
Barber FA. What is the terrible triad? Arthroscop;- 1992: 8:19-22. Leland W. Svstat : the system for statistics. Evanston, IL : Systat Inc.
1988.
DeHaven KE. Diagnosis of acute knee injuries with hemarthrosis. Am
JSports Med 1980: 8:9-14. Liljedahl S-O, Lindvall N, Wetterfors J. Early diagnosis and treatment
of acute ruptures of the anterior cruciate ligament : a clinical and
Farquharson-Roberts MA, Osborne AH. Partial rupture of the anterior arthrographic study of forty-eight cases. J Bone Joint Surg [Am]
cruciate ligament of the knee. J Bone Joint Surg [Br] 1983, 65- 1965; 47-A:l503-13.
B :32-4.
Lysholm J, Gillquist J, Liljedahl SO. Arthroscopy in the early diagnosis
Feagm JA, Abbott HG, Rokous JR. The isolated tear of the anterior ofinjuries to the kneejoint. Acta Orthop Scand 1981 : 52:1 1 1-8.
cruciate ligament. J Bone Joint Surg [Am] 1972 ; 54-A :1340-1.
Mariani PP, Puddu P, Ferretti A. Hemarthrosis treated by aspiration
Galway RD, Beaupr#{233} A, MacIntosh DL. Pivot shift : a clinical sign of and casting: how to condemn the knee. Am J Sports Med 1982;
symptomatic anterior cruciate ligament insufficiency. J Bone Joint I0:343-5.
Surg[Br] 1972: 54-B :763-4.
Noyes FR, Matthews DS, Mooar PA, Grood ES. The symptomatic
Gillquist J, Hagberg G, Oretorp N. Arthroscopy in acute injuries of the anterior cruciate-deficient knee. Part II. The results of rehabilita-
kneejoint. Acta Orthop Scand 1977; 48:190-6. tion, activity modification and counselling in functional disability.
J Bone Joint Surg [Am] 1 983 : 65-A : 163-74.
Glinz W, Segantini P, K#{228}gi
P. Arthroscopy in acute trauma of the knee
joint. Endoscopy 1980: 12:269-74. Noyes FR, Bassett RW, Grood ES, Butler DL. Arthroscopy in acute
traumatic hemarthrosis of the knee : incidence of anterior cruciate
Harilainen A, Myllynen P, Antila H, Seitsalo S. The significance of
tears and other injuries. J Bone Joint Surg [Am] 1980; 62-A:
arthroscopy and examination under anaesthesia in the diagnosis
687-95.
of fresh injury haemarthrosis of the knee joint. Injury 1988: 19:
2 1-4. Noyes FR, Mooar LA, Moorman CT III, McGinniss GH. Partial tears
ofthe anterior cruciate ligament : progression to complete ligament
Irvine GB, Glasgow MMS. The natural history of the meniscus in
deficiency. J Bone Joint Surg [Br] I 989 ; 71 -B :825-33.
anterior cruciate insufficiency : arthroscopic analysis. J Bone Joint
Surg[Br] 1992; 74-B :403-5. O’Connor RL. Arthroscopy in the diagnosis and treatment of acute
ligament injuries of the knee. J Bone Joint Surg [Am] 1974; 56-
Jackson RW, Dandy DJ. Arthroscopvofthe knee. New York, etc : Grune A :333-7.
and Stratton, 1976.
Rand JA. The role of arthroscopy in the management of knee injuries
Jam AS, Swanson AJG, Murdoch G. Hemarthrosis of the knee joint. in the athlete. Mayo C/in Proc 1984; 59 :77-82.
Injury 1983; 15:178-81.
Simonsen 0, Jensen J, Mouritsen P, Lauritzen J. The accuracy of
King J. Indications for surgery in the anterior cruciate deficient knee. clinical examination of injury of the knee joint. Injury 1984; 16:
IntJOrthop Trauma 1991 ; I :138-40. 96-101.
King JB, Aitken M. Treatment of the torn anterior cruciate ligament. Torg JS, Conrad W, Kalen V. Clinical diagnosis of anterior cruciate
SportsMed 1988: 5:203-8. ligament instability in the athlete. Am J Sports Med 1976 ; 4:84-93.

VOL. 75-B, No. 6, NOVEMBER 1993