SympoSium on Endocrinology and diabEtES Inpatient glucocorticoid use: beneficence vs non-maleficence Data suggest that over 12% of inpatients are taking high-dose glucocorticoids, which are a well-recognized cause of hyperglycaemia. Hyperglycaemia is associated with poor outcomes in most medical and surgical specialities, yet rates of blood glucose monitoring and appropriate management remain very low. lucocorticoidsarewidelyusedinseveralspecial- ofthedrug.Shouldpatientsbecounselledonthepoten- ties for their immunosuppressive and anti- tialharmsofthedrug?Beneficencevsnon-maleficence? Ginflammatory properties. However, their use is limited by their well-recognized side-effect profile, Differences between glucocorticoids includingthedevelopmentofhyperglycaemia. Different glucocorticoids have different potencies, as Thevastmajorityofglucocorticoiduseisintheoutpa- showninTable 1.Whilethemajorityofglucocorticoid tientpopulation.Epidemiologicaldatahaveshownthatin useisinmedicalspecialities,thesesteroidsarealsocom- theUKapproximately0.75%ofthepopulationisonoral monly used by anaesthetists as part of the enhanced glucocorticoidtreatmentatanyonetime.Thisvariesfrom recovery after surgery scheme to prevent postoperative about 0.2% in under 30-year-olds to 2.5% in those nausea and vomiting. However, it is not known how between70and79yearsold(RoyalCollegeofPhysicians, manyindividualsundergoingsurgerywhoarenotprevi- 2002).Ofthislong-termuse,40%isusedforrespiratory ouslyknowntohavediabetesbutwhoweregivenreason- disease,withmostoftherestbeingusedforeithermuscu- ablyhighdosesofglucocorticoidhavetheirpre-orpost- loskeletalordermatologicalconditionsthatrequireimmu- operativebloodglucoselevelsmeasured,despiteevidence nosuppression(RoyalCollegeofPhysicians,2002).While that postoperative hyperglycaemia is associated with the vast majority of glucocorticoid use is for less than harm(Frischetal,2010).Recentlypublisheddatashow 5days,overafifthisforover6monthswithabout4.3% that the prevalence of inpatients on supraphysiological beingreportedlyusedforover5years(Fardetetal,2011). glucocorticoids was over 12%, yet only about 20% of Itiswellrecognizedthatlong-termglucocorticoiduse thoseonglucocorticoidswerehavingtheirglucosemoni- is associated with the development of hyperglycaemia, tored(Swafeetal,2014). andobservationaldataformany–ifnotmost–medical andsurgicalconditionssuggestthattheadditionalpres- Pathophysiology of glucocorticoid-induced ence of hyperglycaemia or diabetes is associated with hyperglycaemia pooreroutcomes(Bakeretal,2006;Kwonetal,2013). High-dose glucocorticoids are used predominantly for Thusthequestionarisesthatifsomanyspecialtiespre- theiranti-inflammatoryandimmunosuppressiveeffects. scribe these potentially harmful drugs, should the pre- Howtheyworkisnotcompletelyunderstood,butitis scribers–astheyshouldwitheveryotherdrugtheypre- thoughtthattheyactonglucocorticoidreceptorsinthe scribe–takeresponsibilityforthepotentialconsequences cytoplasmandnucleustosuppresstheexpressionofpro- inflammatorygenesandincreasetheexpressionofanti- Table 1. Relative potency and half-life of commonly used inflammatory genes, which in combination reduce the glucocorticoids production of both arachidonic acid and prostaglandin pathway(FernandesandMcKay,2013). Glucocorticoid Potency (dose equivalent) Duration of action (half-life in hours) Glucocorticoids have several detrimental effects on Hydrocortisone 20 mg 8 carbohydratemetabolism.Theypromotevisceraladipose tissuedepositionandatthesametimeenhancelipolysis. Prednisolone 5 mg 16–36 Theyalterthelevelsofadiposetissue-derivedcytokines Methylprednisolone 4 mg 18–40 and acutely increase hepatic glucose production. They Dexamethasone 0.75 mg 36–54 also have very complex effects on beta cell function (SaltielandKahn,2001;Lambillotteetal,2008). Betamethasone 0.75 mg 36–54 Inthelongertermtheydiminishtheabilityofinsulin Ltd toinitiateintracellularsignallingmechanismsintheliver, Dr Ketan Dhatariya is Consultant in Diabetes, Endocrinology and General adipose tissue and muscle and thereby induce insulin Healthcare Medicine in the Elsie Bertram Diabetes Centre, Norfolk and Norwich University resistance, which in turn inhibits glucose uptake into MA Hospital NHS Foundation Trust, Norwich, Norfolk NR4 7UY muscles and reduces oxidative phosphorylation 2014 ([email protected]) (Hollingdaletal,2008;Petersonsetal,2013). © 252 BritishJournalofHospitalMedicine,May2014,Vol75,No5 British Journal of Hospital Medicine.Downloaded from magonlinelibrary.com by Ketan Dhatariya on June 12, 2014. For personal use only. No other uses without permission. All rights reserved. SympoSium on Endocrinology and diabEtES These effects have the result that, in the very earliest Sulphonylureasarewidelyused,butthereislittlepub- manifestation,ispostprandialhyperglycaemia(Schackeet lished evidence for these. The Joint British Diabetes al,2002).Insomebodywithapreviousdiagnosisofdia- Societieshavewrittenaguidelineforthemanagementof beteswhodevelopshyperglycaemiaasaresultofglucocor- inpatient steroid-induced hyperglycaemia which is due ticoiduse,thiswouldbetermedglucocorticoid-induced to be published shortly and will be freely accessible at hyperglycaemia.However,ifsomebodyisnotpreviously www.diabetologists-abcd.org.uk/JBDS/JBDS.htm. knowntohavediabetesbutdevelopshyperglycaemia,this Severalhospitalswereinvitedtosubmittheirguidelines would be termed glucocorticoid-induced diabetes. The for the management of this condition to the writing post-prandialhyperglycaemiamayprogressandthehyper- groupfortheseguidelinesandalmostallofthemused glycaemiamayjustcauseatransientriseinbloodglucose sulphonylureasfirstorsecondline.However,veryfewof levelsormayresultinhyperosmolarhyperglycaemicsyn- them gave any evidence for their use or references for drome. The best predictors of glucocorticoid-induced this. Figures 1–3 show the pathways for people with diabetesareafamilyhistoryofdiabetesorincreasingage andpreviousglucocorticoiduse(Clementetal,2004). Figure 1. Management of steroid (glucocorticoid)-induced diabetes. Lack of knowledge? No known diabetes ThevastmajorityofinpatientcareintheUKisdelivered • Check random glucose before starting on steroids to identify patient with new onset hyperglycaemia byjuniordoctors.Worklookingatlevelsofjuniordoc- tors’knowledgeaboutdiabetes,andassessingtheirlevels • If the capillary blood glucose is below 12 mmol/litre consider the patient to be at low risk and record the capillary blood glucose daily post breakfast or post lunch ofconfidenceinmanagingtheconditioninthisgroup, hasshownthatmostjuniormedicalstafflackevenabasic • If the capillary blood glucose is found to be greater than 12 mmol/litre the frequency of testing should be increased to 4 times a day grasp of diabetes, and the majority have a poor under- standingofitsmanagement(Georgeetal,2011).Inaddi- • If the capillary blood glucose is found to be consistently greater than 12 mmol/litre (i.e. on two tion,itispossiblethatseniormedicalstaffwhoroutinely occasions during a 24-hour period), then the patient should enter the treatment algorithm below prescribe high-dose glucocorticoids, e.g. those working inrheumatology,renalmedicine,oncology,haematology, Capillary blood glucose readings above desired target (6–10 mmol/litre; acceptable range gastroenterologyandrespiratorymedicine,maynotfocus 4–12 mmol/litre) add in gliclazide 40 mg with breakfast and increase the dose by 40 mg ontheglycaemiceffectsofthisclassofdrug.Itmaywell increments if targets are not reached be that this combination of junior doctors and senior doctorsbeingunawareofthispotentialeffectofglucocor- If no symptoms of hypoglycaemia are experienced by the patient despite being on 160 mg of ticoidtherapymayleadtoglucosenotbeingmeasured. gliclazide in the morning, consider titration to 240 mg in the morning (You may like to seek specialist advice on dose titration at this stage) Treatment options Previousworkhasshownthathyperglycaemiainmedical patientsisassociatedwithincreasedriskofdeath(Baker If still no improvement on maximum dosage consider adding an evening dose of gliclazide or add etal,2006).Therehavepreviouslybeenattemptstotry morning human NPH insulin, e.g. Humulin I, Insulatard or Insuman Basal tofindtheoptimaltreatmentforsteroid-inducedhyper- glycaemia.Theinitialmanagementwouldbeprevention – to try to educate the teams who routinely prescribe Discharge: monitoring will need to be continued in patients remaining on glucocorticoids post discharge high-dose glucocorticoids in helping to detect steroid- • If steroid treatment is ceased in hospital and hyperglycaemia has resolved capillary blood induced hyperglycaemia early and to initiate treatment glucose can be discontinued post discharge whenitisfirstfound.Inaddition,therehasbeensome • If steroids are discontinued before discharge and hyperglycaemia persists then continue with worklookingatavarietyofdifferentdrugsusedtotreat monitoring until normal glycaemia returns or until a definitive test for diabetes is undertaken thiscondition,butthereiscurrentlynoconsensus. (fasting blood glucose, oral glucose tolerance test or glycated haemoglobin) Thiazolidinediones work well in this condition and thereisacomplexinteractionbetweenglucocorticoidsin If steroids are reduced or discontinued: the PPAR (peroxisome proliferator-activated receptor) • Blood glucose monitoring may need to be continued in inpatients and, in discharged patients signallingpathwaythatareoftenthetherapeutictargets assessed by their GP forthisclassofagents(Willietal,2002).However,these