Replacement therapy for invasive procedures in patients with haemophilia: literature review, European survey and recommendations


Cedric Hermans, MD, MRCP (UK), PhD, Haemostasis and Thrombosis Unit, Division of Haematology, Cliniques Universitaires Saint-Luc, Avenue Hippocrate, 10, 1200 Brussels, Belgium.
Tel.: +32 2 764 1740; fax: +32 2 764 8959;


Summary.  Although most surgical and invasive procedures can be performed safely in patients with haemophilia, the optimal level and duration of replacement therapy required to prevent bleeding complications have not been established conclusively. For providing more insight into optimal therapy during invasive procedures, a literature review of surgical procedures in patients with haemophilia was conducted. Concomitantly, current practice was surveyed in 26 European Haemophilia Comprehensive Care Centres, representing 15 different countries. The review identified 110 original papers published between 1965 and 2007. Of these, only two studies were randomized controlled trials. Target levels and the duration of replacement therapy in the published studies were as follows. For major orthopaedic surgery: preoperative targets were 80–90%; postoperative targets showed a high degree of variation, with trough levels ranging from 20% to 80%, duration 10–14 days; for liver biopsy, 70–100%, 1–7 days; tonsillectomy: 90–100%, 5–11 days; indwelling venous access device insertion: 100%, 3–10 days; circumcision: 50–60%, 2–4 days; dental surgery: 30–50%, single treatment. With the exception of dental surgery, current practice in Europe, as assessed by the survey, was largely in agreement with published data. In conclusion, this study provides both a comprehensive review and a large survey of replacement therapy in patients with haemophilia undergoing invasive procedures; these data have informed the consensus practical treatment recommendations made in this paper. This study highlights the need for better-designed studies in order to better define minimal haemostatic levels of replacement therapy and optimal treatment duration.


Most surgical and invasive procedures can be performed safely in patients with haemophilia with factor replacement therapy. Country-specific consensus recommendations for substitutive therapy during these procedures have previously been published [1–8]. However, the optimal level and duration of replacement therapy required to prevent bleeding complications have not been established conclusively. This gives rise to a widely acknowledged lack of consistency among treatment regimens for factor replacement therapy in patients with haemophilia undergoing invasive procedures.

Despite the existence of a large number of reports of the use of replacement therapy cover during invasive procedures in patients with haemophilia, few comprehensive reviews of published literature have so far been performed [9]. Moreover, little data are available regarding the current management of invasive procedures, especially those employing the more recently introduced treatment practices such as continuous infusion or thromboprophylaxis in patients undergoing major orthopaedic surgery.

For these reasons, an extensive literature review of invasive procedures in patients with haemophilia A and B was conducted and a survey carried out to provide information on current practice in 26 European haemophilia centres representing 15 different countries. All survey participants were members of the European Haemophilia Therapy Standardisation Board (EHTSB), an established group of experienced haemophilia centre-based physicians who are responsible for treating a total of 3633 people with severe haemophilia. The aim was to provide a comprehensive overview of existing data that would be invaluable in assessing current practices and identifying areas of controversy and unresolved issues as well as topics for future research. Data that accrued from the literature and the survey, as well as the clinical experience of the treaters, served as a platform for extensive discussions within the network of the EHTSB to develop consensus recommendations for replacement therapy.

Materials and methods

Literature review

A comprehensive review was performed of published evidence regarding replacement therapy for invasive surgical procedures in patients with haemophilia. This included major procedures (orthopaedic surgery including synovectomy) and minor surgery [tonsillectomy, central venous access device (CVAD) insertion, circumcision, dental surgery, liver biopsy]. For each procedure, the literature review was conducted by physicians of the EHTSB. Relevant papers were identified using a Pub-Med search and data were reviewed using a standard protocol. The following information was collected: study reference, year of publication, year of study, level of evidence (rated as 1: high-quality meta analyses, systematic reviews of randomized controlled trials (RCTs), or RCTs with a very low risk of bias; 2: high-quality systematic reviews of case control or cohort studies. High-quality case control or cohort studies with a very low risk of confounding or bias with a high probability that the relationship is casual; 3: non-analytical studies, e.g. case reports, case series and 4: expert opinion), sample size, severity and type of haemophilia, study design (pre-registration, post registration, observational, case-control, retro- or prospective, single centre and multiple centre), type of concentrate (plasma-derived, recombinant, degree of purity and brand), dosage (U kg−1), level of factor aimed for, level of factor achieved, mode of administration (continuous infusion vs. bolus infusion), duration of treatment, complications, outcome and special comments. A critical analysis of relevant papers was then undertaken.


Treatment practices were surveyed by questionnaires related to clinical cases illustrating common invasive procedures, including major orthopaedic surgery (knee replacement), liver biopsy (performed by the transjugular route), circumcision, CVAD (Port-A-Cath®; Smiths Medical International Limited, Kent, UK) insertion, tonsillectomy and dental extraction in patients with severe haemophilia. Information was collected regarding target levels of clotting factors, duration of treatment, treatment modality [continuous infusion vs. bolus, use of DDAVP (deamino d-arginine vasopressin or desmopressin)], peri-operative management including pharmacokinetic evaluation, use of a central line, monitoring of factor levels, use of antifibrinolytics and thromboprophylaxis.

The total number of haemophilia patients followed up by the participating centres was 3633. The average number of haemophilia patients per centre was 241 (range 45–613). The average number of procedures per year per centre was between 2 and 5 for major orthopaedic surgery and liver biopsy, between 6 and 10 for minor surgery and more than 10 for dental surgery. Most respondents were from Comprehensive Care Centres and followed written guidelines for replacement therapy during surgery.


Major surgery

Literature review.  Thirty-five clinical studies were reviewed [10–44] (Table 1). Eight studies were considered to have reached level of evidence 2 and 27 studies up to level 3. Nine studies were multicentre and six were case–control studies with at least historical controls. In total, 1114 patients underwent 1328 major surgical procedures (707 orthopaedic surgeries) including patients with severe, moderate and mild haemophilia A (862) and haemophilia B (241). Surgery was performed using different types of plasma-derived and recombinant factor concentrates. In nine studies, tranexamic acid was used as adjunct therapy. Twenty-three studies used bolus infusion and 16 studies continuous infusion (five studies used both regimens). A small proportion of major surgical procedures (218/1328 or 16.4%) were managed by continuous infusion.

Table 1.   Major surgery in patients with haemophilia: literature review of replacement therapy.
First authorYearReferencesLevel of evidence All (n) A (n) B (n) Major surgery all (n) Orthopaedic surgery (n) Bolus infusion (n) Continuous infusion (n)Factor level, pre-operative (%)Factor level, 1st week postoperative (%)Factor level, 2nd week postoperative (%)Duration of treatment (days) Antifibrinolytics (yes/no) Outcome bleeds (n) Phlebitis (n)
  1. CI, continuous infusion; hc, historical controls; mc, multi-centre; NA, not available; sc, single centre; uc, uncontrolled.

Nilson IM1977103 (sc, uc)776116108531080>90>30–40>10–2014–28Yes40
Krieger JN1977113 (sc, uc)31256581858010060405–12No30
Rudowski WJ1981123 (sc, uc)1018516121331210>50>50NANANA11NA
Willert HG1983133 (sc, uc)181621818180>605050NANo00
Kasper CK1985143 (sc, uc)16316303501943500>80505014NA720
Brown B1986153 (sc, uc)2218423023010050257–14NA40
Kitchens CS1986163 (sc, uc)3630636NA360>80NANA5–18NA21
Martinowitz U1992173 (sc, hc)2525025NA1114>80>50>307–14Yes00
Schulman S1994183 (sc, hc)121201210012>80>50>304–18Yes05
Bushan V1994193 (sc, uc)37325261426080/50–8020–40/   15–3020–40/15–3010No7NA
Lofqvist T1996203 (sc, uc)6653139898980100>30–40>10–2014–28Yes10
Hay CR1996213 (sc, uc)24240212002110080NA5 (CI)No00
Shapiro AD1997222 (mc, uc)740743424340>60>30NA10No00
White GC1997232 (mc, uc)1313095NANANANANANANA0NA
Srivastava A1998243 (sc, uc)181172014200>80/>60>20–40/  >15–30>15–30/>10–2011No10
Heeg M1998253 (sc, uc)9811212120>100>50>2514No10
Campbell PJ1998263 (sc, hc)211801818810100>80NA13–17No81
Gosh J1998273 (sc, uc)161247270>60>30NA10Yes20
Negrier C1998283 (sc, uc)13941310013>80>80>509–22No00
Tagariello G1999293 (sc, hc)15141119011>80>70/>40>40/>2010Yes02
Rochat C1999303 (sc, hc)5505505>80>50NA5 (CI)No05
Scharrer I2000312 (mc, uc)151508480NANANA12No00
Batorova A2000322 (sc, c)4040043311825>80>50>3012Yes34
Bastounis E2000333 (sc, uc)654315586580>80>30>3014No20
Chowdary P2001343 (sc, uc)6065305>80>80NA3–10No02
Scharrer I2002352 (mc, uc)22220137130NANANA12–26No00
Mishra V2002363 (sc, uc)9628808>90>50–70>309Yes00
Ragni MV2002372 (mc, uc)260262311149>80NANA10–20No01
Dingli D2002383 (sc, uc)282803525035>80>80>506 (CI)No50
Hoots WK2003392 (mc, uc)280282521025>90>70NA6 (CI)No03
Evans G2003403 (sc, uc)4045505>90>70NA3–54No10
Lusher JM2003412 (mc, uc)4242048NA480>70NANANANA00
Wolf DM2004423 (mc, uc)8805550>90NANA9–21NA00
Stieltjes N2004433 (mc, uc)161601815018NANANA5–21Yes41
Lee V2004443 (sc, uc)9819972>80>30>10–205–44No00
 111486224113287071101218     13125

Factor levels were not available in four of the studies. Considering the other 31 studies, the preoperative target factor levels for haemophilia A and B were >90% in 11 studies, >80% in 15 and >50–70% in five. Thus the majority of studies (26/31) aimed at values >80%. In the 27 studies that addressed postoperative trough levels for the first week, values were >70% in eight studies, >50% in 11 and >20–30% in eight. For the second week, data were available in 18 studies and levels were >50% in four, >30% in seven and >10–20% in seven.

Duration of treatment in the 31 studies varied from 5 to 14 days in 19 of the studies, 15–21 days in six and up to 28 days or longer in further six studies. Bleeding complications occurred in 131/1328 or 10% of major surgical procedures; most of them (96/131 or 73.3%) were in the 714 procedures included in the papers published before 1990. During the above time period, there were two cases of fatality related to bleeding complications.

No reliable correlation between postoperative bleeding and target factor level emerged. Local phlebitis was observed in a limited number of patients treated with continuous infusion (25/218). No publication recommended thromboprophylaxis with anticoagulants.

Survey.  A target factor level of at least 80% was reported by all centres prior to major orthopaedic surgery (Table 2) which is similar to published practice. In contrast to the details reported in the literature, according to which continuous infusion was employed in only 16% of the procedures, this infusion technique was used in procedures performed in nearly half of the centres. In most centres, factor levels were maintained above 80% in the postoperative period. With bolus infusion, two infusions per day were administered in order to maintain trough levels above 80% during the immediate postoperative period (from day 1 to 5) and around 60% in the late postoperative period (from day 6 to 14) – values that are higher than published targets. The duration of postoperative replacement therapy ranged from 12 to 14 days. When continuous infusion was used, the factor level was maintained at 80% during the first five postoperative days and decreased to 30–40% or 50–60% between days 6 and 14. Before major orthopaedic surgery, pre-operative pharmacokinetic evaluation was performed in one-third of the centres and the recovery measured in more than half of them. Thromboprophylaxis with low-molecular-weight heparin is used in more than half, and antifibrinolytic treatment in more than two-thirds, of the centres.

Table 2.   European survey of replacement therapy for invasive procedures in haemophilia.
ProcedureDuration of treatment (days)Replacement therapy
Target FVIII levels in %
No treatment (%)
  1. CVAD, central venous access device; CI, continuous infusion.

 Pre-operative 8119 0
>7 694
CVAD insertion
 Pre-operative 8119 0
 Postoperative1–32562.50 12.50
>7  694
 Pre-operative 100  0
 Postoperative1–36931 0
4–72575 0
>7 252550
Surgical synovectomy
 Pre-operative 87.5012.50 0
 Postoperative1–35050 0
>7 502525
Liver biopsy
 Pre-operative 771490
Dental extraction
 Pre-operative 325990
Knee arthroplasty
 Pre-operative 100   
 Postoperative1–5 (bolus)851500
6–14 (bolus)771220
1–5 (CI)663400
6–14 (CI)058420


Literature review.  Four papers reporting data on replacement therapy for synovectomy published between 2000 and 2003 were identified [45–48]. The level of evidence was 3 or 4. These four studies involved 158 patients undergoing 197 different procedures. Factor replacement always included a loading dose, ranging from 15 to 50 IU kg−1 in patients with haemophilia A and from 30 to 90 IU kg−1 in patients with haemophilia B, aiming at factor levels between 30–100% and 30–90%, respectively. Subsequent replacement therapy was either short (repeated bolus at full dose 8–12 h later and at half-dose on day 2) or prolonged for 8 weeks using a prophylactic regimen (20 IU kg−1 three times per week – 2 weeks and 15 IU kg−1 two times per week – 6 weeks for haemophilia A and 30 IU kg−1 three times per week – 2 weeks and 25 IU kg−1 two times per week – 6 weeks for haemophilia B). No bleeding complications were reported.

Survey.  A target level of factor VIII (FVIII) between 80% and 100% was reported by 87.5% of the centres (Table 2). Continuous infusion for this procedure was considered as an option by 62.5% of the treaters. Treatment was continued for more than 7 days in a majority of the centres. Antifibrinolytics were used in 56% of the centres.

Liver biopsy

Literature review.  A total of 26 papers published between 1977 and 2007 were reviewed [49–74] (Table 3). These studies included 778 procedures performed in 713 patients (Haemophilia A: 526; Haemophilia B: 68; unspecified: 119). Among these patients, 278 had severe haemophilia. Although the review criteria did not include patients with inhibitors, one study reported the inclusion of one patient with inhibitors. Patients co-infected with HIV were not excluded unless they were severely immunosuppressed. The following types of procedures were performed: percutaneous liver biopsy (11 studies), biopsy under laparoscopy (two studies) and transjugular liver biopsy (seven studies). The replacement therapy protocol did not vary in most studies with a pre biopsy target level of FVIII of 100% in most studies. Intermittent bolus infusions every 12 h were usually given after the procedure. Continuous infusion was also used in some patients (four studies). The post biopsy target levels were 70–100% for the first 24 h, 50–70% for 24–48 h, and >50% for 48–72 h. When the replacement therapy was extended for 5–7 days, target levels were >30%. The duration of in-hospital stay ranged from 4 h to 7 days. For patients discharged early, treatment was continued at home. Concomitant use of antifibrinolytics was reported in two studies. All studies except one reported bleeding risk similar to non-haemophilic patients (0.5%). The incidence of complications and the factor requirements did not differ among the various methods.

Table 3.   Liver biopsy in patients with haemophilia: literature review of replacement therapy.
First authorYearReferencesLevel of evidenceDesignNumber of patientsHAHBNumber with severe diseaseNumber of proceduresMethod of liver biopsyFactor level prebiopsy (%)Factor level post biopsy and duration of replacementUse of antifibrinolyticsOutcome
  1. CI, continuous infusion; HA, haemophilia A; HB, haemophilia B; NA, not available; TJLB, transjugular liver biopsy.

Lesesne HR1977493sc66  6Percutaneous10072 hNoNA
Preston FE1978503sc88 88Percutaneous10072 hNoNA
White GC1982513sc1515 1515NANS72 hNoNA
Aledort LM198552 mc115  90126NANSNANoTwo deaths
Hay CRM1987533sc343222443NANANANoNo bleed
Makris M1991543sc7766114399NANANANoNo bleed
Ahmed MM1996553sc5050  50Percutaneous10048 hNoNo bleed (pain, 2)
Hanley JP1996563sc2222  22Laparoscopy100 (A) 70 (B)50–100% (48 h) (HA)/50–70% (48 h) (HB)NoNo bleed
Wong VS1997573sc3535  35Percutaneous10050% (36 h) – CI in 5 patients/2–4 days in hospitalNANo bleed
Gupta R1997583sc65156Transjugular80–10024 h in hospitalNoNo bleed
Fukuda Y1998593sc36285 36PercutaneousNA50% loading dose (24–48 h)NoNo bleed
Adamowicz A1999603sc13   13NA100 (A) 80 (B)100% (HA) 80% (HB) (12–24 h)/50% (24–48 h)NoNo bleed
Farell RJ1999613sc55  5Percutaneous100CI (4 U Kg−1 h−1) (48 h)NoNo bleed
McMahon C2000623sc171341021Percutaneous100100% (48–72 h)4 procNo bleed
Venkataramani A2000633sc1293512Percutaneous100>50% (24–48 h)/CI >30% (7 days)NoBleed (1)
Shields PL2000643sc2121   NANANANANo bleed
Lethagen S2001653sc272431139Percutaneous100100% (48 h)/2 days in hospitalYesNo bleed
Delladetsima J2002663sc2424  25Percutaneous100NANANo bleed
Denzer U2003673sc11   Mini Laparoscopy NANANA
Dimichele DM2003683sc1091810Transjugular>70>70% (24 h)/>50% (24–72 h)/>30% (4–5 days)NoNo bleed (pain, 3)
Stieltjes N2004693sc6960214588TransjugularNS20 U kg−1 per 8–12 h (48 h) (HA)/60 U kg−1 per 8–12 h (48 h) (HB)NoBleed (4)
Saab S2004703sc1192711Transjugular10072 hNoNo bleed
Shin J2005713mc56479765TJLB (64), Femoral (1)75%half-dose 12–24–48 hNoBleed (7)
Dawson MA2005723sc54115Transjugular100%100% (24 h) – >50% (48 h)NoNo bleed
Detrait M2007733sc99069Transjugular100%80–100% (24 h) then 50%NoNo bleed
Sterling RK2007743sc29245 29Percutaneous100%50–30% (24 h)–50% (24–72 h) (HA)
50% (during 72 h) (HB)
NoNo bleed
Total71352668285778    12 bleeds

Survey.  The target factor level before transjugular biopsy was between 80% and 100% in 77% of the centres (Table 2). Bolus infusions were used in two-third of the participants. Post biopsy, the factor level was maintained above 80% in 45% of the centres, or between 40% and 70% in 45%. There was wide variation among the centres, although the duration of treatment was usually between 2 and 3 days in 10 out of the 18 centres.

Paediatric surgery

Three invasive procedures (tonsillectomy, circumcision and CVAD insertion) commonly performed in children with haemophilia were addressed in the survey.


Literature review.  Only three studies involving tonsillectomy published between 1985 and 1996 were identified [75–77] (Table 4). All were single-centre, retrospective studies and only included 24 patients. The level of evidence was 3. The target factor level preoperatively varied between 80% and 100%. Replacement therapy was maintained for 5–11 days in combination with antifibrinolytics. Only one patient developed a bleeding episode postoperatively.

Table 4.   Tonsillectomy and circumcision in patients with haemophilia: literature review of replacement therapy.
ProcedureFirst authorYearReferencesLevel of evidenceNumber of patientsSeverity of haemophiliaStudy designType of concentrate Dose (IU kg−1) Factor level aimed at (%)Factor level achieved (%)Duration of treatment (days)Outcome (early outcome)Special comments
  1. C, controlled; CI, continuous infusion; FG, fibrin glue; NA, not available; SC, single centre; R, retrospective; O, observational; P, prospective; TA, tranexamic acid.

TonsillectomyThach T19857539Mild, moderate, severeSC, RNA40–5080 initially then 30–50NA9–11No complication 
TonsillectomyPrinsley P19937635Mild, moderate, severeSC, RNANA90–100 then >5090–1005–71 epistaxis at day 9 (FVIII level at 40%)Use of TA until day 7
TonsillectomyConlon B199677310Mild, moderate, severeSC, RNANA90–10090–10010No complicationUse of TA until day 7
CircumcisionMartinowitz U199278310Severe onlySC, R, 0NANANANANA3 secondary bleeds (2 required replacement therapy), FG usefulUse of TA and FG. No replacement therapy
CircumcisionKavakli K19977934Severe onlySC, R, OBranded Plasma FVIII productsNA50–6050–604No complicationUse of TA and FG
CircumcisionAvanoglu A199980222NAC, P bolus or CI (4 U Kg−1 h−1)NANA2–4 Reduction of substitutive therapy by adjunction of FG
CircumcisionShittu OB200181370Mild, moderate, severeSC, R, ONoneNANANANA52.1% bleeds 
CircumcisionZulfikar B200382356Mild, moderate, severeSC, R, ONA20–30NA30–40 (1–4 days)/20–30 (5–7 days)/10–20 (8–12 days) No major bleed, five transient bleeds, one delayed haematomaProlonged replacement therapy with low dose. Use of TA
CircumcisionKaraman MI200483345Mild, moderate, severePNA25–40NANA7–18Five minimal bleeds 

Survey.  In agreement with the published data, all participants would aim for a factor level between 80% and 100% (Table 2). Interestingly, 33% of participants would consider the use of continuous infusion. Treatment would be continued for 1–3 days in 69% of the centres aiming at levels of 80–100% and for 4–7 days in 75% of centres. Fifty per cent of treaters would however consider replacement therapy of more than 7 days duration. Antifibrinolytics were used in 94% of the centres.


Literature review.  Six studies involving circumcision published between 1992 and 2004 were reviewed [78–83] (Table 4). Four were retrospective and two prospective; the level of evidence was 3 in five studies and 2 in one. These studies included 163 patients. Replacement therapy with factor concentrate was used in five studies. In one study, antifibrinolytics were used alone. The factor level before the procedure was only reported in two studies where it was between 30% and 60%. Adjuvant haemostatic agents (fibrin glue or antifibrinolytics) were used in most studies. Postoperatively, replacement therapy was continued for 2–8 days. Bleeding complications were reported postoperatively in three studies with a frequency up to 50%.

Survey.  In 81% of the centres, the target factor level before circumcision was between 80% and 100% (Table 2). Use of continuous infusion was considered a treatment option by 12.5% of the respondents. Treatment was continued for 1–3 days in 75% of the centres aiming at levels of 50% on average and for 4–7 days in 44% centres. Only 66% of the treaters consider longer treatment to be appropriate. Antifibrinolytics were used in 69% of the centres.

Port-A-Cath® insertion

Literature review.  Fifteen studies involving Port-A-Cath® insertion and published between 1992 and 2004 were reviewed [84–98] (Table 5). Eleven studies were retrospective. The level of evidence was 3 in 11 studies and 2 in two studies. The studies involved 256 patients who underwent 347 procedures. Seventy patients had inhibitors. For patients treated with FVIII or FIX concentrates, the target level before surgery reported in eight studies was more than 90%. Replacement therapy was maintained for a variable period ranging from 2 to 10 days. Bleeding complications were reported following 34 procedures.

Table 5.   Port-A-Cath® insertion in patients with haemophilia: literature review of replacement therapy.
First authorYearReferenceLevel of evidenceSample Size (number of patients and severity)Study designType of concentrate Dose (IU  kg−1) Level aimed at (%) Level achieved (%)Duration of treatment (days)Outcome (early outcome)Special comments
  1. APCC, activated prothrombin complex concentrate; INH, inhibitor; ITI, immunotolerance; NA, not available; O, observational; P, prospective; PAC, Port-A-Cath; R, retrospective; sc, single centre; TA, tranexamic acid; uc, uncontrolled.

Ljung R199284312 severe haemophilia AO, R, SCNA40–50NANANATwo mild haematomasUse of TA
Girvan DP19948539 severe for 11 procedures (4 PAC vs. 7 external devices)O, R, SC NA90–1001007Two cases of minor bruising 
Liesner RJ199586323 severe (6 INH)/27 proceduresR, 2 centresHigh purity. INH (6): 5 PorcFVIII, 1 rFVIIa45–108 >755–727% complication rate (6/23 with 3 bleeds, 2 new inhibitors, 1 allergy) 
Blanchette VS199687319 (1 moderate and 18 severe, 3 INH)/23 PACO, R, SCPorcFVIII or aPCC for INHNA10090–1007No complication13 prophylaxis/2 ITI/4 other indications
Smith OP19968833 severe haemophilia A INH/4 proceduresO, R, SCrFVIIa90 μg kg−1NANA2No complicationUse of TA. Ports represented 2 out of the 4 procedures
Perkins JL199789332 severe (7 INH)/36 proceduresO, R, SCNANA100NA5–10No complication 
Warrier I199790322 severe (11 INH)/35 proceduresO, R, SCPorcFVIII or rFVIIa for INH50–10080–90NA5–7Mild bleeds in five patients (14% of procedures)9 cases for ITI
Santagostino E199891215 moderate and severe (2 INH)O, P, UC, SCRecombinant FVIII or rFVIIa70–8080–9080–9061 case with haematoma at day 7 and inhibitor at day 14Indications for PAC: prophylaxis 13 cases, ITI 2 cases
Miller K199892345 (3 moderate, 42 severe) (8 INH)/49 procedures (41 PAC)O, R, 2 centersNANA100>30–403–10No major complication. 11 minor haematomas (22% of procedures)Indications for PAC: prophylaxis 26 cases, ITI 8 cases
Van Den Berg HM1998934 Review  100NA5–7NA 
Montoro JB19989431 severe hemophilia A (INH)case reportrFVIIa (Continuous Infusion)90 μg kg−1  5Mild haematoma at the site of insertionUse of TA.
Bollard CM200095314 severe (5 INH)/23 proceduresO, R, SCVarious50–100NANA3–5Four haematomas (17% of procedures)Indications for PAC: prophylaxis 11 cases, ITI 2 cases
McMahon C200096346 severe (8 INH)/77 procedures (74 PAC)RNA NANANANAIndications for PAC: prophylaxis 34 cases, ITI 12 cases
Morado M200197315 (moderate and severe with INH)/34 proceduresO, R, SCaPCC or rFVIIa100 U kg−1 (aPCC)/120 μg kg−1 (rFVIIa)NANANA2 Haematomas out of 6 portsOnly 6 ports
Valentino LA2004982256 patients/347 procedures 70 INHReview  100 3–8  

Survey.  Target FVIII level was between 80% and 100% in 81% of the centres (Table 2). Use of continuous infusion was considered an option by 12.5% of the treaters. Treatment was continued for 1–3 days in 62.5% of the centres aiming at levels of 50% on average and for 4–7 days in 31% of the centres. Antifibrinolytics were used in 69% of the centres.

Dental surgery

Literature review.  Twenty-one studies involving dental surgery published between 1965 and 2006 were identified [99–119] (Table 6). The level of evidence was 1 in two studies and one subgroup in another study. Ten studies were level 2 evidence and nine were level 3.

Table 6.   Dental extraction in patients with haemophilia: literature review of replacement therapy.
First authorYearReferencesLevel of evidenceNumber of patientsProceduresStudy design% of procedures on severe patientsStudy armsBleeds in each armHaemostatic treatmentComments
  1. CC, case–control; DB, double blind; EACA, epsilon-aminocaproic acid; F, factor; FG, fibrin glue; MC, multi-centre; O, observational; P, prospective; R, retrospective; RCT, randomized controlled trial; SC, single centre; TA, tranexamic acid.

Part 1
Biggs R19659933843SC, O, R85Arm 1: single extraction, 5 days of treatment0/8 bleedsFHigh bleed rates despite prolonged treatment
Arm 2: 2–9 extractions, 5–10 days of treatment9/13 bleeds
Arm 3: >9 extractions, 8–12 days of treatment 7/13 bleeds
Tavenner RWH196810023061SC, CC, R47No factor replacement. EACA 6 g 4-hourly pre-operative until discharge FG and suture7/30 bleeds (6/7 with severe disease)EACA (s), FGFG, EACA and suture effective in mild disease
Walsh PN197110121818SC, CC, R53Arm 1: factor replacement target level 50% during 7–10 days + EACA 6 g qds during 10 days7/18 bleedsF, EACA (s) 
Arm 2: factor replacement target level 50% during 7–10 days7/20 bleeds
10113131MC, RCT39Arm 1: factor replacement target level 50% + EACA (s) 6 g qds during 7–10 days3/15 bleeds F, EACA (s)EACA is effective (level 1 evidence)
Arm 2: factor replacement target level 50%14/16 bleeds
Forbes CD197210212832SC, RCT, DB47Arm 1: plasma + TA 1 g tds during 5 days2/16 bleedsPlasma, TA (s)TA is effective (level 1 evidence)
Arm 2: plasma11/16 bleeds
Tavenner RWH197210322151SC, CC, R41No factor replacement. TA (s) 1.5 g qds pre-operative until discharge. FG and suture5/51 bleeds (5 with severe disease)FG, TA (s)Fibrin glue, TA (s) and suture effective in mild disease
Ramstrom G197510425297SC, CC, R40Arm 1: factor replacement target level 30–50% pre-operative, 25–30 during 2–3 days, 12–25% during 6–10 days23/27(9/23 prolonged or severe)F, TA (s)Antifibrinolytics and local haemostasis are of benefit
Arm 2: similar to arm 1 except many patients received a single pre-operative bolus to 20–30%. TA 1 g tds (duration not stated) 10/30 (0/10 severe or prolonged)
       Arm 3: factor replacement target level 8–10% pre-operative + TA (s) 1 g tds during 7 days + local haemostatics (surgical/acrylic plates)4/27 (0/4 severe or prolonged)
Kaneda T198110537498SC, O, RNAFactor replacement until haemostasis achieved12% (3.5–25%) for haemostasisF Factor level of 25% may be adequate
Suzuki M19831063913SC, O, R70Thrombin and packing. No factor replacement therapy.3/13 bleedsFGLocal thrombin generation is important
Steinberg SE198410731619SC, O, R44In 11 patients, FVIII increased to 20% and FIX to 10% pre-operative. 8 patients (4 severe) had no pre-operative treatment. Thrombin was given when necessary.7/19 bleedsFHigh bleed rate in absence of antifibrinolytic therapy, FG and factor concentrate.
Baudo F198510832929SC, O, R45FG, collagen and suture8/29 (7/13 with severe disease)FGFG and suture may prevent bleeding in mild disease but bleeding rate is high in severe disease.
Stajcic Z198510924343SC, CC51Arm 1: factor replacement with 10–20 U kg−1 FVIII + EACA (s) 6 g qsd during 6–10 days0/13 bleedsF, EACAAntifibrinolytics are of benefit
Arm 2: factor replacement as for arm 1 (pre-operative) + FVIII (? Dose) (postoperative) + EACA (locally) during 1 day5/11 bleeds
Arm 3: factor replacement as for arm 1 (pre-operative) + EACA (locally) during day 1 + EACA (s) 6 g qds during 6–10 days5/11 bleeds
Part 2
Sindet-Pedersen S198611024054SC, O, R62Arm 1: factor replacement target level 60–70% pre-operative + TA (s) 37–123 mg kg−1 per day during 3–6 days. Some post treatment was given as some patients were on prophylaxis13/27F, TA (s and locally)TA mouth wash appears to add benefit to TA (s).Factor level of 10% preop level is inadequate in severe disease.
Arm 2: as for arm 1 + TA (mouth wash) during 5–7 days0/12 bleeds
Arm 3: Factor replacement with median preop level of 13% (10–22%). If baseline factor level >10%, no treatment plus TA (s) 72–106 mg kg−1 per day during >6 days + TA (mouth wash) during 5 days at least6/15 bleeds
Baudo F198811125991SC, O, R60Arm 1: FG11/63 bleedsFG, TA (s)Antifibrinolytics are of benefit
Arm 2: FG + TA (s) 1 g tds during 7 days0/28 bleeds
Ramtsrom G1989112, 104287228SC, O41Arm 1: Factor replacement target level 8–10% pre-operative + TA (s) 1 g tds during 7 days plus local haemostasis, acrylic plates or suturing of flap.5/162 bleedsF, TA 
Arm 2: Similar but time period 1972–1973 and reported in arm 3 in reference 104
Rakocz M199311323737SC, CC, R100Arm 1: FG9/12 bleedsFG, TA (locally)Antifibrinolytics are of benefit
Arm 2: FG + TA (mouth wash) qds during 10 days3/25 bleeds
Waly NG199511412424SC Arm 1: replacement therapy + TA (s)75% bleedsF, TA (locally)TA mouth wash effective
Arm 2: replacement therapy + TA (s) + TA (mouth wash)8.4% bleeds
Zanon E20001152261261SC, CC, R45Arm 1: factor replacement target level 30% pre-operative, TA (s) 20 mg kg−1 pre-operative then tds during 7 days, FG and suture2/77 bleedsF, FG, TA (s)FG and 30% factor rise is effective.
Arm 2: standard extraction in healthy controls. 117/184 had suture1/184 bleeds
Piot B200211634551SC, O, R20Factor level 50% pre-operative and 24 h postoperative and TA (s) 20 mg kg−1 tds during 8 days0/51F, TA (s)50% factor rise pre-operative and 24 h postoperative + TA (s) 8 days is effective
Frachon X200511731619SC, O, R36Factor replacement preop with 50 IU kg−1 FVIII in patients with severe haemophilia A. DDAVP in patients with mild haemophilia A. Local haemostasis (FG, suturing). TA locally for 3 days6/19 bleedsF, FG, TA (locally)Single treatment with factor not effective in all cases
Franchini M20051183135139MC, O, R38Factor replacement or DDAVP aiming at 50% pre-operative with systemic and local antifibrinolytics7/139 bleedsF, FG, TA (s) and locallyFactor replacement plus antifibrinolytics is safe and effective
Correa ME200611933131SC, O, R25FG, EACA (s) during 7 days6/31 bleedsFG, EACA (s) 
  11241470 989    

Several studies were randomized. A total of 1124 patients underwent 1470 surgical dental procedures. Clotting factor was used in 15 studies, fibrin glue in 10 studies and antifibrinolytic agents in 17 studies. The factor level before extraction was 50% in most studies. The duration of replacement therapy varied between 5 and 7 days.

Survey.  Factor VIII was given in all patients in order to raise the FVIII level to between 60% and 80%. Treatment was repeated in one-third of the centres. Antifibrinolytics were given to all patients for a period ranging from 5 to 10 days.


Although numerous reports of the use of replacement therapy with clotting factor concentrates in patients with haemophilia undergoing different kinds of surgical procedures have been published, no comprehensive review of the literature has so far been performed. It is noticeable that the most recently published international treatment recommendations and guidelines of replacement therapy in haemophilia are not supported by a comprehensive literature review. This article aims to address this gap in the literature by providing an original and comprehensive literature review as well as a survey of current practice among a large group of European haemophilia treaters.

Interesting conclusions can be drawn from the literature review. It is evident that there is a lack of robust data, as, with the exception of two studies in dental surgery, no randomized controlled trials of replacement therapy were identified. For the remaining studies reviewed, the level of evidence was rated as 2, but more frequently 3. Thus, the body of evidence we have is founded on studies, most of which have significant methodological limitations. Procedures such as tonsillectomy and circumcision are commonly performed yet are poorly represented in the literature; only nine of the 110 papers considered these operations. Additional limitations are the very small patient numbers included in the majority of studies and the fact that many reports do not provide detailed information regarding the levels of clotting factor achieved or the duration of replacement. The percentage of bleeding complications and the magnitude of blood losses are not systematically reported so that minimal haemostatic levels cannot be clearly derived from our review of the literature.

With the exception of a single study performed in India [24], there has been no attempt during the last three decades to test the haemostatic efficacy of using lower factor levels. This observation can to some extent be accounted for by the fact that most studies were performed in developed countries (Europe and USA) where the availability of clotting factor concentrates is greater and where there has been a tendency to aim for higher target levels and to use higher doses of factor concentrates in surgical procedures over the last few decades.

The survey conducted among a large group of treaters caring for several thousands of haemophilia patients has provided invaluable information on treatment practices including target factor levels, duration of treatment and use of certain treatment modalities such as continuous infusion and thromboprophylaxis in the postoperative course of major orthopaedic surgery. The survey highlights much heterogeneity in practice consistent with the wide range of treatment regimens reported in the literature. However, it was interesting that for most procedures reviewed, there was a good agreement between published data and current treatment practices in terms of intensity and duration of replacement therapy. For dental care, however, the prescribed treatment regimens revealed by the survey are usually more intensive and target higher factor levels than those reported in the literature. Continuous infusion appears to be used in nearly half of the patients undergoing major orthopaedic surgery in the survey centres, suggesting that the extent of the clinical experience and use of this treatment modality is greater than what the published literature would suggest.

Because of the limitations discussed and the very limited number of randomized controlled trials, no well-supported recommendations can be made regarding optimal factor levels and duration of replacement therapy for patients with haemophilia undergoing invasive procedures. However, data accrued in the literature and the survey, as well as the clinical expertise of the treaters involved in the group, have served as a platform for extensive discussion within the network of the EHTSB and the development of consensus recommendations for replacement therapy. Consensus was reached on the following recommendations and the level of evidence for each of these is given in parenthesis (Table 7):

Table 7.   Grading of recommendations and levels of evidence.
GradeLevelType of evidence
AIaEvidence obtained from meta-analysis of randomized studies
AIbEvidence obtained from at least one randomized controlled trial
BIIaEvidence obtained from at least one well designed controlled study without randomization
BIIbEvidence obtained from at least one other type of well-designed quasi-experimental study
BIIIEvidence obtained from well-designed non-experimental descriptive studies, such as comparative studies, correlation studies and case-control studies
CIVEvidence obtained from expert committee reports or opinions and/or clinical experience of respected authorities
  • 1In patients undergoing major orthopaedic surgery, including open surgical synovectomy, preoperative factor levels should be 80–100% (grade B, level III). In the postoperative period, minimal factor levels should be maintained above 50% in the first postoperative week and above 30% in the late postoperative period (grade C, level IV). Continuous infusion appears safe and effective and the use of antifibrinolytic agents and thromboprophylaxis could be considered in certain settings.
  • 2In patients undergoing liver biopsy, the preoperative factor level should be above 80% and replacement therapy should be continued for at least 3 days (grade B, level III). The biopsy method should be selected depending on the local experience.
  • 3Replacement therapy is required for children undergoing surgical procedures such as tonsillectomy, CVAD insertion and circumcision (grade B, level III). The preoperative factor level should be above 80% and replacement therapy should be continued for 7–10 days after tonsillectomy and at least 3 days after CVAD insertion (grade B, level III). For circumcision, a target level of 80% and continuation of replacement therapy during 3–4 days are recommended (grade C, level IV). Adjunctive treatment with antifibrinolytics and/or fibrin glue should be considered (grade B, level IV).
  • 4For patients undergoing dental extraction, replacement with clotting concentrate is recommended with a minimal factor level of 50% (grade B, level IV). Antifibrinolytic therapy is recommended for 7 days (grade A, level I). Adjunctive treatment with fibrin glue should be considered (grade B, level IV).

In conclusion, this study provides both a comprehensive review of the available literature and a large survey of replacement therapy in patients with haemophilia undergoing invasive procedures; this review highlights the need for robust future studies on the incidence, type and magnitude of bleeding complications in patients with haemophilia undergoing invasive procedures and comparing them with bleeding events in non-haemophilic patients. Within the constraints discussed, treatment recommendations are provided, which reflect the literature, current practice and the clinical experience of the EHTSB. Better-designed studies using standardized protocols, as well as patient registries, are however needed in order to define minimal haemostatic levels and optimal durations of treatment.


The European Therapy Standardisation Board consists of the following members and European haemophilia centres: Alessandro Gringeri, Milan, Italy; Ana Villar and Marta Morado, Madrid, Spain; Angelika Batorova, Bratislava, Slovakia; Angiola Rocino, Napoli, Italy; Anastasia Karafoulidou, Athens, Greece; Barry White, Dublin, Ireland; Carmen Altisent, Barcelona, Spain; Cedric Hermans, Brussels, Belgium; Chantal Rothschild, Paris, France; H. Marijke van den Berg, Utrecht, the Netherlands; Géraldine Lavigne Lissalde, Montpellier, France; Gerry Dolan, Nottingham, UK; Herve Chambost, Marseille, France; Jan Astermark, Malmö, Sweden; Jerzy Windyga, Warsaw, Poland; Lorenzo Giovanni Mantovani, Napoli, Italy; Manuela Fraga, Porto, Portugal; Mario Schiavoni, Bari, Italy; Mario von Depka, Hannover, Germany; Michael Richards, Leeds, UK; Philippe de Moerloose, Geneva, Switzerland; Rosario Pérez Garrido, Seville, Spain; Riitta Lassila, Helsinki, Finland; Robert Klamroth, Berlin, Germany; Thierry Lambert, Paris, France.


Barry White has received an unrestricted educational grant from Baxter and NovoNordisk. The other authors stated that they had no interests which might be perceived as posing a conflict or bias.