Preventing severe infection after splenectomy

BMJ 2005; 331 doi: http://dx.doi.org/10.1136/bmj.331.7514.417 (Published 18 August 2005) Cite this as: BMJ 2005;331:417
  1. Adrian Newland, professor (a.c.newland{at}qmul.ac.uk),
  2. Drew Provan, senior lecturer,
  3. Steven Myint, editor in chief
  1. Department of Haematology, St Bartholomew's and the Royal London School of Medicine and Dentistry, London El IBB
  2. Department of Haematology, St Bartholomew's and the Royal London School of Medicine and Dentistry, London El IBB
  3. Journal of Infectious Disease, The Brae, Dunmow, Essex CM6 1HU

    Patients should know the risks, be immunised, and take prophylactic antibiotics

    Splenectomy may be followed by severe systemic infection because such surgery removes the splenic macrophages that filter and phagocytose bacteria and other bloodborne pathogens. Overwhelming post-splenectomy infection (OPSI), as this complication is called, is uncommon but has high mortality.

    Overwhelming post-splenectomy infection is usually caused by the encapsulated bacteria Streptococcus pneumoniae, Haemophilus influenzae, and Neisseria meningitidis and more than half of those infected die.1 Other pathogens in such infection may include bacteria such as Escherichia coli and Pseudomonas aeruginosa,2 Capnocytophagia canimorsus, group B streptococci, Enterococcus spp, Ehrlichia spp, and protozoa such as the Plasmodium spp leading to malaria.

    The first description of overwhelming post-splenectomy infection was published by King and Schumaker in 1952.3 The disease may begin as a minor flu-like illness that rapidly escalates into a fulminant infection.2 It is most common in the first two years after splenectomy but may occur decades later.4 Its true incidence is not known, but the estimated incident rate among patients who have had splenectomy is 0.18-0.42% per year, with a lifetime risk of 5%. There are only limited data relating to morbidity and mortality, primarily because of a lack of systematic studies.

    Bisharat et al reviewed all 78 studies published between 1966 and 1996.5 Twenty eight contained data relating to incidence, morbidity, and mortality and the effects of infection in different age groups. Of 19 680 patients who had had a splenectomy, 3.2% developed invasive infection, and the overall mortality was 1.4%. The mean interval between splenectomy and infection was 22.6 months. The incidence of infection was higher for patients with thalassaemia major (8.2%) and sickle cell anaemia (7.3%) than for patients with idiopathic thrombocytopenia (2.1%), and higher in children with thalassaemia major (11.6%) and sickle cell anaemia (8.9%) than in adults with the same diseases (7.4% and 6.4%, respectively). The most reliable incidence data are probably those of Schwartz et al, who found a risk of fulminant infection of one case per 500 person years of observation.

    The main risk factors are the age at which splenectomy occurs, with children being particularly at risk; the reason for splenectomy; and the time interval from splenectomy (most cases occur within two years).6 The reliability of studies on the incidence has been hampered by limited statistical precision and by study groups that are too small with wide variation in age ranges of the patients studied, and by heterogeneity in the underlying pathology leading to splenectomy and the indications for splenectomy.

    Guidelines published by the British Committee for Standards in Haematology emphasised that most infections after splenectomy could be avoided through measures that include offering patients appropriate and timely immunisation, antibiotic prophylaxis, education, and prompt treatment of infection.7 These guidelines have been updated8 and now recommend vaccination at least two weeks before splenectomy, oras soon as possible after emergency splenectomy has been performed—for example, after trauma.

    Vaccines should include those for pneumococcal infections, with boosters every five to 10 years as dictated by levels of antibody titres; and for H influenzae type B, against which most children in the United Kingdom will have been immunised already: the effect of reimmunisation against this pathogen is not known. Previous recommendations did not include meningococcal vaccination,7 but the revised guideline recommends meningococcal serogroup C conjugated vaccine for unimmunised patients.8 Patients should also be offered annual influenza vaccination. Despite these guidelines, a recent audit in Scotland showed that immunisation practice for patients after splenectomy varied considerably, with only 13% of patients receiving all three vaccines before elective splenectomy.9

    The role and efficacy of antibacterial prophylaxis remains unclear, and not all countries recommend it. Some studies show that patients may not adhere to prophylaxis but, in the absence of robust systematic evidence against this approach, the chief medical officer in the United Kingdom recommends lifelong prophylaxis with oral phenoxymethylpenicillin (or erythromycin for people who are allergic to penicillin). Patients who develop infection despite vaccination and antibacterial prophylaxis should go into hospital for treatment with broad spectrum systemic antibiotics. In addition, patients might keep a supply of antibiotics at home, changing from prophylactic to therapeutic doses if they develop febrile illness. This is particularly important for those who do not, or will not, take prophylaxis, a group whom doctors should make more aware of the risk and dangers of infection.

    Patients need to know the nature and likelihood of overwhelming post-splenectomy infection and that they should seek medical attention if they become ill and feverish. The audit by Kyaw et al showed that only half of patients received information about antibiotic prophylaxis.9 Each patient should carry at all times a letter or card documenting the splenectomy and should wear a bracelet or pendant to signal their status to doctors and other healthcare staff. Healthcare records in hospital and primary care should clearly record that these patients have had splenectomy. Primary care doctors, who may be responsible for keeping patients fully immunised, should know what vaccinations these patients have had.


    • Competing interests None declared


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.