Use of calcium antagonists and need for perioperative transfusion in older patients with hip fracture: observational study

BMJ 1997; 314 doi: (Published 01 March 1997) Cite this as: BMJ 1997;314:643
  1. Giuseppe ZuccalÀ, assistant professora,
  2. Marco Pahor, assistant professorb,
  3. Francesco Landi, assistant professora,
  4. Giorgio Gasparini, assistant professorc,
  5. Francesco Pagano, research fellowa,
  6. PierUgo Carbonin, professora,
  7. Alberto Cocchi, associate professora
  1. a Department of Internal Medicine and Geriatrics, Catholic University of the Sacred Heart, 00168 Rome, Italy
  2. b Department of Preventive Medicine, University of Tennessee, Memphis, USA
  3. c Department of Orthopaedics, Catholic University of the Sacred Heart
  1. Correspondence to: Dr ZuccalÀ
  • Accepted 12 November 1996


Calcium antagonists inhibit platelet aggregation,1 which may increase the risk of bleeding.2 Whether these drugs increase postsurgical haemorrhage remains controversial,3 but more attention is being given to reducing surgical bleeding because of the risks associated with homologous transfusions.4 We evaluated the haemorrhagic potential of calcium antagonists in hip surgery, which is associated with considerable transfusion requirements.

Patients, methods, and results

We prospectively evaluated the need for perioperative transfusion in 161 consecutive elderly patients with hip fracture (mean (SE) age 80 (0.5) years; 82% women) who had cemented total hip arthroplasty (85 subjects), hemiarthroplasty (16), or osteosynthesis (60) in 1994-5. One hundred patients had hypertension, 19 diabetes, 47 vascular disease, six liver disease, and 17 peptic disease. Nitrates were taken by 56 patients, angiotensin converting enzyme inhibitors by 44, digitalis by 39, diuretics by 45, and corticosteroids by nine. Seventy patients took calcium antagonists (50 nifedipine; 14 amlodipine; 6 nimodipine) at home and throughout their hospital stay.

All patients received diazepam and atropine as premedicants. Anaesthesia was induced with thiopentone and vecuronium and maintained with nitrous oxide, isoflurane, and fentanyl. Intraoperative arterial blood pressure was maintained within 20% of preinduction values. Crystalloid and colloid were infused at a standardised rate. Packed cell volume was determined at 30 minute intervals during surgery; haemoglobin concentrations were determined preoperatively, at the end of surgery, the next morning, and after one week. Allogenic packed red blood cells were administered as required in single units, either intraoperatively to maintain a minimum packed cell volume of 28% or postoperatively to achieve a minimum haemoglobin concentration of 95 g/l.4 No patients received scavenged red cells. Subcutaneous heparin and intravenous ketorolac were started on admission.

Fig 1
Fig 1

Number of transfused blood units according to use of calcium antagonists. P value was calculated by the Kruskal-Wallis H test. Full squares represent the mean number of transfused blood units

Transfusions were required in 52 (74%) patients taking calcium antagonists, but only 30 (33%) of the remaining subjects (P<0.0001; fig 1); the mean (SE) number of units received were 1.7 (0.22) and 0.9 (0.10) respectively (P<0.0001). Among patients who did not receive transfusions, the fall in haemoglobin concentration after surgery was greater in patients taking calcium antagonists than in other patients. (24.2 (3) v 13.6 (1.5) g/l after 24 hours and 33.5 (3.2) v 18.6 (1.7) g/l after one week, P<0.005). Patients taking calcium antagonists showed increased prevalence of hypertension (90% (63) v 41% (37); P<0.001) and vascular disease (43% (30) v 19% (17); P<0.002) and higher blood pressure (152 (2) v 142 (2) mm Hg; P<0.002) than the other patients. Other demographic, clinical, laboratory (including clotting), and surgical parameters showed no variation with use of calcium antagonists.

After hypertension, vascular diseases, and other variables associated with the outcome (age, sex, preoperative haemoglobin concentration, and type and duration of surgery) were adjusted for, Cox regression analysis gave the risk of transfusion associated with calcium antagonists as 2.05 (95% confidence interval 1.14 to 3.70). The population attributable risk of receiving transfusions for calcium antagonists was 24%. When only patients with hypertension (n=100) were included the adjusted relative risk for calcium antagonists was 2.35 (1.15 to 4.78). No dose related gradient was found. When specific drugs were examined separately similar risks were found for those taking nifedipine (relative risk=2.67; 95% confidence interval 1.24 to 5.75), and amlodipine (2.36; 1.30 to 4.28) as in non-users. Too few patients took nimodipine for analysis.


Our findings indicate that calcium antagonists are associated with increased perioperative transfusion requirements and postoperative haemoglobin loss. Calcium antagonists may have enhanced bleeding in our patients directly or through some interaction with ketorolac.1 5 If an interaction existed this may explain the lack of a dose dependent effect.2 5 Many strategies have been proposed to reduce surgical transfusion requirements and ensuing morbidity.4 Our results suggest that switching patients treated with calcium antagonists to alternative drugs before surgery might reduce by 24% the number of patients requiring transfusions.



  • Funding None.

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract