Therapeutics overview

Prescribing in general – the essential medicines list concept.

The Essential Medicines List is not intended to contain every medication. The lists have been set internationally to include adequate therapy for 80-90% of illnesses.

Accessing the EML is very feasible even though the app may not be functional; as soon as chapters in any of the lists are updated, the latest versions appear here:

Is it possible to obtain agents not on the list? Yes, but with sound written motivation accompanied by some good supporting evidence.

What if agents on the list aren’t available? Check with your local pharmacy or regional pharmacist whether they should be available at that site.

Patient characteristics – weight, age, liver and renal function

Weight. In order to prescribe using the weight, you have to weigh the patient. This is not always possible – moving a moribund morbidly obese patient to scale is unwise.

Age. Older patients have mildly impaired renal and hepatic function.

Liver and kidney. In general in the elderly or those with liver or kidney disease:

• If a safer alternative exists, consider using it.
• With chronic medication, start with low doses – there’s usually no rush.
• Consider medication concordance (adherence, compliance)

Influencing factors:

• Doses per day – fewer is probably better.¹
• Tolerability of individual pills (size, adverse reactions, interactions).
• Understanding of the reasons for taking that specific treatment.
• Level of belief in the efficacy of treatment.
• Degree of discomfort caused by the underlying condition.
• Suitability of the regimen relative to life style.
• Availability of medication – prescribing medication that is unavailable at a clinic will clearly lead to non-compliance with that agent.

What you can do:

Make the regimen as simple as possible.

Try to prescribe fewer than five medications per day, and preferably all once daily.

Discuss therapy as often as needed – patient understanding and ownership of a regimen is valuable.

Give preference to better tolerated treatments. A slightly less effective medication taken regularly works better than a more effective one that isn’t taken at all.

Prescription checklist

A prescription is the culmination of quite a lot of hard work on your part and deserves that little extra care and attention. Most of us will occasionally make a mistake when writing a script, but the more you try to check off the items in the table each time you prescribe, the fewer your errors will be.

Appropriate doses and dose ranges.

Most texts give dose ranges rather than precise figures, or else they give doses in mg per kilogram or per square metre. Until one becomes familiar with an individual drug, it is often difficult to know what dose to prescribe.

Many doses are derived from pharmacokinetic work on healthy volunteers who are not obese. For purposes of calculation, use ideal body mass.

Look up vial and tablet sizes – it doesn’t help to prescribe a precisely calculated dose that is technically difficult to dispense, as it will inevitably be rounded off.

For drugs with nervous system activity (antiepileptics, psychotropics, antidepressants, etc) start at the lower end of the range and work up.

Be particularly careful of overdosing elderly folk with suspected renal impairment. Don’t forget the option of increasing the dose interval under such circumstances.

Prescribing antibiotics

There are two common sources of confusion – insecurity about correct dosing from the dose ranges given in formularies, and lack of awareness of the difference between an in vitro minimum inhibitory concentration (MIC) and an in vivo situation incorporating distribution, penetration, immunity and infection severity. Resolution of both issues can be aided by remembering a few points.²

There are 3 broad mechanisms of antibiotic action:

• time dependent killing (penicillins, cephalosporins, macrolides and clindamycin) where effect is proportional to duration of time above MIC,
• concentration dependent killing (aminoglycosides, metronidazole and amphotericin) where effect is proportional to concentration reached, (so doses should be high but can be less frequent),
• time dependent with persistent effects as well (tetracyclines, fluconazole, vancomycin) where the total amount given is important (area under the curve or AUC)

Persistent effects. Antibiotics may still exert an effect even after levels have fallen below the MIC. This is commonly called the ‘post-antibiotic effect’.

Synergism e.g. Pseudomonas and Proteus infections should be treated with two antibiotics from different classes (an aminoglycoside and a third generation cephalosporin) although fourth generation cephalosporins or carbapenems are adequate as monotherapy for these organisms.

For severe infections, use doses at the higher end of the ranges suggested in formularies, rather than in the middle of the range – e.g. penicillin G 5 MU 6 hourly for infective endocarditis but only 2 MU 6 hourly for pneumonia.

Antibiotic duration – give until clinical improvement except where there are guidelines supporting fixed duration (e.g. at least six weeks for lung abscess or infective endocarditis, fourteen days for pneumococcal meningitis.) In patients with uncomplicated pneumonia, changing to oral treatment after three days, assuming good clinical response at that time, gave the same outcomes as continuing IV therapy for a further five days.³

Antibiotic stewardship

Antibiotics are a precious resource in the fight against infections, and there are growing concerns about resistance. A systematic review⁴ looked at the odds of developing resistance after being prescribed an antibiotic found the odds were nearly doubled in the months following a script, but remained elevated for up to a year after treatment (e.g. OR for resistance to the particular antibiotic after treatment of a chest infection 2.4, 95% CI 1.4 to 3.9. at two months, and still 2.4, but with a wider confidence interval (1.3 to 4.5) at one year.)

Antibiotic stewardship is not just about selecting the best drug and optimising dosing and duration of use but also about recognising that as clinicians we have a societal responsibility to future patients not to act irresponsibly with the treatment of current patients. This means that biggest is not always best, and it is imperative to ‘think lean’ – what is the best way, in terms of minimising future resistance, that I can treat this patient safely?

An educational intervention in primary care in the UK⁵ (Wales) demonstrated that it is feasible to change practice, although the size of the changes was modest (4.4 % reduction in prescribing of antibiotics, 5.5% reduction in costs, no increase in hospitalisations or re-visits with the same complaint.)

Stewardship is not just about personal prescribing habits, but about networking monitoring, and changing the mindset of prescribing communities (both patients and clinicians.)

Perspective – aminoglycoside toxicity

Do not prescribe an aminoglycoside to any patient with renal impairment if a safe and equally effective non-nephrotoxic alternative is available. The use of aminoglycosides in the elderly is particularly problematic – one prospective cohort study⁶ identified features independently associated with the development of nephrotoxicity in 249 elderly (mean age 75) patients:

1. Hb<10g/100ml or pH < 7.38.
2. In hospital for more than 7 days before aminoglycoside started
3. Use of aminoglycoside for more than 11 days
4. A high aminoglycoside trough level.
5. Concomitant diuretic use.

In this group of patients, 12.4% developed > 50% rise in creatinine, but in only 6% did the level rise above 186 micromol/l. No patients developed oliguric renal failure, and in all cases the creatinine fell prior to discharge. Thus in patients with an initial creatinine clearance of more than 30ml/min the short term use of an aminoglycoside (4-5 days) is reasonably safe, but caution should be exercised when one or more of the risk factors is present. In the South African context, a new concern is the combination of an aminoglycoside for MDR TB and tenofovir, and if a patient needs the former, then TDR should be stopped and zidovudine (if HB normal) or stavudine (if anaemic) used instead.

Perspective – dosing of ceftriaxone

Ceftriaxone may interact with calcium containing solutions (e.g. Ringer’s), and so the manufacturer recommends that at least 48 hours elapse between using the two. A further issue concerns doses and dosing intervals for this agent. In general 1g daily is adequate. In bacterial meningitis use 2g 12 hourly, if it is needed for endocarditis, then use 2g daily. There have been reports of increased clearance in critically ill patients with severe sepsis⁷, and in such situations it may be worth increasing to 1g 12 hourly. In some areas, pneumococcal MICs for ceftriaxone would support using a higher dose for community acquired pneumonia: 2 d daily.

Fluid therapy

Assessing hydration status

There is a difference between ‘diagnosing dehydration’ and ‘confirming normal hydration’ (are you using the signs to ‘rule in’ or ‘rule out’?) Consider:

• A history suggesting significant fluid loss (e.g. vomiting or diarrhoea).
• Signs of intravascular volume depletion
• Pulse rate, and postural change in pulse (>30 bpm).
• Blood pressure and postural change in BP (>20 mmHg).
• Response to a fluid challenge.
• Signs of total body fluid depletion (less reliable)
• Axillary moisture.
• Tongue dryness and grooving.
• Capillary filling time.
• Biochemical features
• Urine specific gravity.
• Serum chemistry.

There is some evidence to suggest that certain signs are of greater value in blood loss hypovolaemia than water loss, and that the performance of some signs may vary with age and nutritional status. Think carefully, and if in any doubt, do serum chemistry.

Perspective – ‘test’ performance when assessing hydration:

A postural rise in pulse by more than 30 bpm has a sensitivity of only 43%, but a specificity of 75% (LR+ 1.7, LR- 0.8), and a postural drop in BP of 20 mmHg is even less sensitive (29%) but marginally more specific at 81% (LR=1.5, LR- 0.9.) Dry axillae were also relatively unhelpful (sens 50%, spec 82%, LR+ 2.8, LR-0.6.) A dry mouth occurred more commonly (sens 85%) but was non-specific (58%) with LR+ of 2.0 and LR- 0.3. Sunken eyes also did not perform well (Sens 62%, spec 82%, LR+ 3.4, LR- 0.5.)⁸

No single sign is very good at determining the absence of dehydration, and nearly all the studies were of suboptimal quality. Unfortunately one suspects that the better performance of the tests for ruling in the diagnosis reflects the presence of patients with such severe dehydration that its clinical determination is not a diagnostic issue.

Perspective – colloid versus crystalloid for fluid resuscitation

This debate has been raging for years, and there are cogent theoretical reasons for the use of crystalloids in some circumstances. A multi-centre study of 537 patients with severe sepsis from Germany⁹ found a statistically significant difference in the rate of development of acute renal failure (34.9% versus 22.8%, p = 0.002) in those given hydroxyethyl starch (HES) rather than Ringer’s lactate. Starch colloids should no longer be used in critically ill ICU patients¹⁰.

Perspective – saline versus Ringer’s or other balanced solutions

A further debate concerns the use of Ringer’s lactate (Na 131, Cl 111 mmol/l) or normal saline (Na 154, Cl 154 mmol/l) in resuscitation of ‘medical’ (non-trauma) patients. The latter has been associated with normal anion gap hyperchloraemic metabolic acidosis when given rapidly in very large quantities and there is some laboratory evidence of coagulopathy with large volume resuscitation, but there still seems to be little high quality evidence of clinically important differences between the two when used in standard ICU situations¹¹ . A systematic review looking at randomised controlled trials and cohort studies¹² failed to demonstrate a mortality advantage of balanced solutions over saline. However another systematic review¹³ including only the RCTs, while also failing to demonstrate any impressive differences, made the case that from a probabilistic perspective, despite no statistical difference by frequentist rules (1% increased mortality ranging to 9% reduction in mortality), there was a high probability (90%) that balanced crystalloids reduced mortality… On an overall 90 day mortality of 22% in the control arm, this would thus range between a 2% absolute reduction and a 0.2% absolute increase in mortality. If the products cost the same, there is probably little reason not to prefer balanced crystalloid.

(There is a major marketing interest in being able to identify a proprietary balanced solution with a survival advantage, as IV fluids are a multi-million dollar industry.)

Administering fluids

• The best oral/nasogastric fluid to maintain fluid balance in the absence of GIT or renal losses is tap water. It is readily available, cheap, patients and staff are familiar with its use, and it has few side effects.
• Salt infusions are difficult for the body to handle. It is far easier for the kidneys to deal with a litre of water or 5% dextrose than with a litre of normal saline.
• 5% dextrose water is ‘acidic’ and certain drugs precipitate in it – e.g. phenytoin. 10% dextrose is sclerosant, so drip re-siting is often needed sooner than expected.
• A fluid prescription is just like any other prescription in terms of clear indications about dose, duration, frequency and concentration.
• Keeping check of what is actually infused can be quite a challenge, but is very important. Just because you have written a fluid prescription does not mean that the fluid has been given…

Blood products

Packed red cells. Indications for packed cell transfusion are symptomatic anaemia due to acute ongoing gastrointestinal blood loss or haemoptysis, and chronic anaemia associated with adequate prognosis but poor marrow reserve (aplastic anaemia, chronic leukaemias, and myelodysplastic syndromes). Transfusing patients with nutritional anaemias is seldom appropriate and may be dangerous. In general, don’t transfuse unless you can see the bleeding or know the cause.

Red cell product nomenclature and ordering.

A conventional packed cell unit is called buffy coat poor red cell concentrate. Order this for most patients. If the pack has been further filtered to remove even more of the antigenic white cells, it is called leukocyte depleted red cell concentrate, and this is appropriate for patients likely to receive multiple transfusion – e.g. aplastic anaemia, chronic leukaemias. Very rarely, washed red cell concentrate is required for patients who have had severe allergic reactions to plasma or have specific anti-IgA antibodies. Bone marrow transplant or severely immunosuppressed patients may occasionally require irradiated RBCs or platelets.

Platelets. Their main role is in patients with thrombocytopenia due to aplastic anaemia or acute leukaemia with a dangerous bleed (CNS, genitourinary or gastrointestinal). Patients with ITP rapidly develop tolerance to infused platelets, which under these circumstances may only remain in the circulation for a matter of minutes.

Fresh frozen plasma. Emergency correction of bleeding in patients over-anticoagulated with warfarin. Infuse 10 ml/kg (usually about 2 packs) and assess response. Also consider in patients with bleeding due to severe liver disease with a raised INR. Supportive role in severe DIC – first try to remove the cause.

Factor VIII. See section on haemophilia.

Albumin. It is of little value in sick ICU patients, with an initial meta-analysis suggesting harm¹⁴ being refuted by a subsequent RCT¹⁵ which didn’t show harm, but also didn’t show any major advantages over saline. It may prevent late hypotension after large volume paracentesis.

Post-operative (and other) analgesia

Pain management is not just about prescribing a drug. It is useful to think about the type of pain that the patient is likely to be experiencing (visceral, musculoskeletal, neuropathic or spasmodic (colic)) and adjust the drug accordingly. Dosage should be dictated not only by size, but also by the patient’s response, as different individuals appear to perceive the same pain differently. Also pay attention to dosing intervals as infrequent dosage leads to fluctuating drug levels and poor pain control.

Special situations:

Post-op, any evidence of chest trauma – ensure regular NSAID until patient can cough freely and says is pain free.

Evidence of cardiac failure or renal impairment – beware of using NSAIDS.

In healthy individuals doses of ibuprofen of up to 1200 mg per day (i.e. 400 mg 3x/d) has no more gastric side effects than placebo¹⁶ . In the short term, doses of twice this (2400 mg/day in divided doses) are approximately equivalent to indomethacin or diclophenac 50 mg 3x/d.

The addition of mild opiates (e.g. codeine or tramadol) to ibuprofen causes a small benefit (e.g. about 8% improvement in pain score)¹⁷ but also increases the adverse effects.

From an orthopaedic perspective, while there is evidence in animal models of a measurable effect of NSAIDS on bone healing, there is little evidence that short term use (days) causes clinically relevant harm.

Perspective – NNTs for various analgesics

An adaptation of one analgesic league table (Bandolier’s Oxford Pain site) appears below. The modification has been made to include only medication commonly available in public service facilities. It applies to acute pain and reflects a surgical bent (e.g. dental pain, bone pain) yet nevertheless gives a useful perspective on some deeply ingrained half-truths.

Opiates.

Codeine. This agent is used for cough suppression and as part of combination analgesics. The real clinical benefit in cough suppression is uncertain, and its efficacy for this purpose in patients with haemoptysis is unproven although widely recommended.

Tramadol. This is another weak opiate which is cheaper than codeine but otherwise quite similar. Dose starts at 50mg 4x/d, increasing to a maximum of 400 mg/day. Evidence for its efficacy in cough suppression is limited to a few case reports.

Morphine. Dose four hourly – reducing to six hourly rarely works well. Dangerous respiratory depression in healthy individuals with genuine severe pain is probably rare – the data about this appears to have been derived from its use in healthy pain-free volunteers. Clearly extra caution should be exercised in the frail and the elderly. Dose 10-15 mg 4 hourly. Remember to manage constipation (lactulose is probably as good as sennosides in cancer patients),¹⁸ and in peri-operative use, where nausea is in any case common, consider adding metoclopramide 10 mg 8 hourly.

Pethidine – the purported advantages in terms of visceral spasm (e.g. cholecystitis) are unconvincing, and morphine is probably acceptable. In adequate doses, there is little difference in analgesic efficacy and a good case can be made for not using it at all.¹⁹ There is weak evidence to suggest that pethidine may be more likely to predispose to arrhythmias and fits in frail and elderly patients, and it may lead to more tachyphylaxis with repeat dosing than morphine.

Other opiates. Although this list may appear very short, all of the other new opiates currently undergoing various degrees of marketing hype, together with the other older agents provide little if any clinical advantage outside of anaesthetics.

Corticosteroid prescribing

General principles

In the absence of liver failure, prednisone and prednisolone are equivalent (prednisone is the pro-drug). Methylprednisolone has slightly less sodium retaining effect and more anti-inflammatory effect but this is probably rarely of clinical significance

Parenteral steroids are more expensive than oral steroids.

Intramuscular ‘depot’ steroid preparations are sometimes erratically absorbed and indications for this route of use are limited. If there are serious questions about potential adherence to oral prednisone after an admission for acute asthma, imtramuscular methylprednisolone may be an adequate alternative²⁰. The dose in one trial²¹ was 160 mg of methylprednisolone.

Complex stepping down regimens are of dubious value and quite difficult to follow – keep it simple, without too many steps.

Dexamethasone is traditionally given in situations where inflammatory oedema reduction is desired (intra-cerebral metastases, tuberculoma with mass effect, etc.) It has little if any sodium retaining effect, but the evidence that it is clinically superior to prednisone is scanty. In situations where oral dexamethasone is difficult to access and hypertension/fluid retention is not a major problem, giving prednisone in a dose 7 times that of dexamethasone is reasonably effective – e.g. 30 mg prednisone 3x/d is equivalent to 4 mg dexamethasone 3x/d. Betamethasone is usually substitutable on a mg for mg basis with dexamethasone.

Indications for corticosteroids.

Adrenal replacement.

Giving hydrocortisone 20 mg mane and 10 mg nocte works for most individuals. In some cases, patients may still complain of tiredness and postural dizziness, in which case adding fludrocortisone 0.05 mg per day is appropriate. There have been some recent suggestions that a three times per day regimen of hydrocortisone may give a biologically smoother profile. Prednisone 5 mg daily is an adequate alternative to hydrocortisone if the latter is unavailable.

Immunosuppression.

If steroids are being used for immunosuppression, it is acceptable to give the whole dose once a day, unless the dose is so high that it is uncomfortable to take this number of tablets all in one go. Most reasons for immunosuppression require a short course of high dose therapy and then a rapid drop either to nothing or to low dose maintenance. Adrenal suppression is only clinically relevant if the steroids have been given for longer than 2 weeks.

A patient on 30-40 mg per day of prednisone for an acute asthma attack can have the course stopped after 10 days, and can be left on maintenance inhaled steroids. Someone with an SLE flare can have a short course and then be tailed to 10 mg per day. There are very few indications for stepping down from 60 mg by 5 mg increments each week. In conditions, such as ITP, where tailing steroids may lead to recurrence, the drops should be slower (e.g. give a 10 mg lower dose each month till reach 20 mg, then drop by 5 mg per month)

Perspective – corticosteroids in head injured patients.

Although this is arguably a surgical topic, the subject illustrates the dangers of adopting therapies in the absence of good evidence of efficacy. Corticosteroids have been given to individuals with head injury for more than 30 years based on early suggestive trials and appealing physiological explanations. A Cochrane review had shown a small but non-significant mortality benefit (1% relative risk reduction) and this led to considerably less enthusiasm for their use, but little evidence to suspect actual harm.

The MRC CRASH trial²² involved 10 000 adult patients and used methylprednisolone 2g load over one hour, then 0.4g per hour for 48 hours. It showed 21.1% mortality at 2 weeks in the corticosteroid arm and only 17.9% in the placebo arm – an absolute increase in mortality of 3.2% (CI 0.019 – 0.045). This can be expressed as a NNH (number needed to harm) of 32 – for every 32 patients given corticosteroids, one died. A review²³ provided the alarming but possibly a little fanciful calculation that globally the use of steroids for head injuries during the 1970s and 1980s may have led to an excess 10 000 deaths…

Immunizations and antibiotic prophylaxis

Hepatitis B vaccination

The main indications for vaccination against hepatitis B are to do with being either a contact (e.g. household member) or a health care worker. Immunise if well and Hepatitis B surface Ab negative. The HBsAb titre used as a cut-off for those requiring a booster (i.e. previously immunised) is often taken as 10 IU/l. Antigenicity of the vaccine is variable, and subcutaneous injections into very fat areas may not ‘take’ – always inject into deltoids, not buttocks.

Influenza vaccination.

A case can be made for giving this to all elderly persons, all health care workers, and any patients with severe chest disease, particularly COPD²⁴ The absolute benefit in COPD is modest, and derived from a small number of patients in old trials, but if in the interval after vaccination, the placebo group were to have one COPD exacerbation per person, those vaccinated would have 0.37 fewer episodes (i.e mean difference 0.37, 95% CI 0.11 to 0.64.)

Its efficacy in the general healthy adult population is less marked, with one review suggesting it reduces the proportion of individuals experiencing influenza in a single season from 2.3% in the unvaccinated to 0.9% in those given the vaccine, i.e. a 1.4% absolute risk reduction, giving a number needed to vaccinate to prevent one event of 72; this halved to 29 if the end point was ‘flu-like illnesses’ rather than confirmed influenza.²⁵ The same review found only low quality evidence regarding days off work prevented by vaccination (0.04 days less, 95%CI 0.14 days less to 0.06 days more. That is a one hour difference.

Pneumococcal vaccination

This is indicated for those about to undergo splenectomy. It use is also advocated in old age home residents in the United States. A systematic review²⁶ failed to demonstrate mortality benefit (OR 0.90, 95% CI 0.74 to 1.09) but found evidence of a reduction of invasive pneumococcal disease (OR 0.26, 95% CI 0.14 to 0.45). This translates into an absolute risk reduction of 0.3% (0.4% reduced to 0.1%), and a NNT of 367, 95% CI 274 to 592.

Meningococcal meningitis prophylaxis

Close contacts (household contacts) of an index case with confirmed or suspected meningoccal meningitis or meningococcaemia should be considered for antibiotic prophylaxis. Clear evidence of efficacy is limited, but drugs suggested are either:

Rifampicin 600 mg 12 hourly for 2 days OR

Ciprofloxacin 500 mg stat.

Perspective – efficacy of prophylaxis for meningococcal meningitis.

One review²⁷ identified five studies addressing this issue, and was able to pool data from only three, all of which were retrospective observational studies. One of the others was a tiny RCT from Florida, which only included 54 patients and had no events in either placebo or intervention groups, making calculation of efficacy somewhat difficult… The fifth study could not be included in the pooled estimate because of dissimilarity in design.

The baseline risk (control patients) of contracting disease in the next month varied from 0.44% to 2.8%, and the pooled risk ratio with prophylaxis was 0.11, giving an ARR of 0.46%. and a NNT of 218 (95% CI 121 to 1135.) This means that although prophylaxis causes an 89% reduction in the risk of an event, because disease in contacts is rare, more than 200 household contacts would need to be treated to prevent one case.

There are no adequate trials allowing for an evidence-based approach to recommendations about what to do about day care contacts – some countries do and others do not offer this.

General points:

• Pregnant women and small children should avoid elective travel to high risk areas
• Most areas contain organisms with varying degrees of chloroquine resistance.
• Barrier protection with lotions containing diethyltoluamide, and impregnated nets and long sleeves is always advisable.
• Mefloquine is currently popular because of its efficacy and convenience – however when there are concerns about the CNS side effects consider alternatives such as doxycycline or atovaquone/proguanil. Of these three options, only doxycycline is readily available in the public service.

Trials of therapy

Reasons suggested for the need for a trial of therapy are:

Therapeutic urgency. There is sometimes a need to start treatment without diagnostic certainty because the patient is so sick that delay is considered dangerous. In this situation, the usual diagnostic process is continued in parallel. Commencing therapy isn’t a signal to stop diagnosing.

Diagnostic languor. Sometimes the prescriber is too tired or too busy to investigate the patient correctly and takes a prescribing guess in the hope that response to therapy will obviate the need for further thought or investigation.

Diagnostic frugality. A more experienced clinician may feel comfortable in dispensing with diagnostic steps that would normally be considered appropriate. Therapeutic response should not then be seen as ‘post hoc’ diagnostic corroboration.

Barriers to investigation. In this case there is a logistic (e.g. power cut in laboratory) or medical (e.g. too ill to transfer to CT scan) impediment to following a conventional diagnostic pathway. Alternatively, it may be difficult to visualise a further step (‘diagnostic exhaustion’?). Treatment response is then an investigative way of choosing between remaining alternative diagnoses; however the mere existence of a prescription does not define a diagnosis.

The steps in undertaking a therapeutic trial involve:

• If possible, first discuss it with somebody else who may think of something you have forgotten and thus avoid the need for the trial.
• Don’t forget to also discuss the rationale with the patient, and make sure he or she understands and is in agreement with you.
• Chose the condition you want to try to treat first. In a generally well patient, one would start with the one requiring the least noxious therapy (primum non nocere) whereas in the very sick patient one would aim for the one that is most immediately dangerous – e.g. Staph. pneumonia before TB.
• Define your treatment modality carefully so that it is as specific as possible for only that condition. (e.g. rifampicin may partially heal a necrotising pneumonia if being used as part of a trial of TB therapy; giving dexamethasone as part of the treatment for a tuberculoma will cause the oedema around a glioma to shrink.)
• Choose your output measures. For example, the cardiothoracic ratio would be a valid marker of response in a trial of therapy for a suspected tuberculous pericarditis. On the other hand using FEV1 in a trial of steroids in COPD would not be appropriate in a patient being discharged after an acute exacerbation. (Lung function tests would be expected to improve anyway in the following few weeks). Check that baseline measures are stable (weigh more than once prior to discharge…)
• Decide in advance when you are going to review the patient and make a judgement about success or failure of the trial.
• Write down in advance your criteria for failure/success of the trial.
• Write down in advance your plan of future management if the trial fails and the patient is not getting better.

Adverse drug reactions

Acute drug reactions

There are a number of reasons why a patient can be acutely unwell or collapse, usually after an injection of a drug:

• Anaphylaxis – hypotensive and a tachycardia, with or without wheeze, skin rash.
• Vagal reaction – hypotensive and a bradycardia. No rash or wheeze.
• Serum sickness – delayed (days) – emphasis on skin and joints rather than CVS.
• Procaine reaction – after procaine penicillin which has inadvertently gone partly or entirely intravenously instead of intramuscularly. The patient feel unwell, and sometimes may become talkative or disinhibited, and then may have fits. A blood pressure drop is not always very marked. There is usually no wheeze or skin rash.
• Known effect of a drug given appropriately but too rapidly – e.g. large doses of intravenous furosemide given rapidly can cause acute deafness; ‘red man’ syndrome can occur after or during rapid infusion of vancomycin.
• Idiosyncratic neurological reactions – e.g. prochlorperazine or metoclopramide may induce acute dystonia.
• Drug-organism interaction leading to an inflammatory response – e.g. Herxheimer’s reaction from penicillin in syphilis.
• The only one of these that needs adrenalin is anaphylaxis. The only ones that are drug ‘allergies’ are anaphylaxis and serum sickness. Probably the commonest cause for being labelled penicillin allergic is a vagal reaction, but it is very important to take a history and examine the patient, including BP and pulse at the time of the reaction.
• Manage anaphylaxis by giving 1 ml of 1:1000 adrenalin S.C or I.M. (deltoid). (The debate about S.C. versus IM is of unclear clinical relevance – in a shocked patient, perfusion of both muscle and skin is poor, so it is not going to be well absorbed from either site.)
• Repeat the adrenalin IV (if still hypotensive) once venous access is available.
• Giving hydrocortisone does nothing in the acute stage (it takes hours to work), and giving an antihistamine in the absence of a major wheeze and itch is of unclear value although commonly recommended.
• Consider salbutamol nebulisation for bronchospasm.
• Don’t forget the 40% facemask oxygen.
• Don’t neglect to document your observations, including serial pulse and BP.
• If a genuine allergy is diagnosed, tell the patient, write it on the outside of the folder (as well as clearly documenting it on the top of prescription sheets), and attempt to organise a patient-carried warning bracelet or necklace.

Subacute and chronic adverse drug reactions.

Drug reactions are a particularly important cause of morbidity, and most can be reversed simply by stopping the offending agent. Common ‘diseases’ caused by drugs are hepatitis, skin rashes, chronic cough and blood disorders (leukopenia and thrombocytopenia). Less well recognised are interstitial nephritis, dehydration (diuretics) and confusional states. They can be thought of as either idiosyncratic or as the side effect of an inappropriate dose.

Other common medications involved are the anti-tuberculous group, antiepileptics (all of them), laxatives, diuretics, and the antihypertensives. The absolute risk (cases of skin rash per million users per week) is 0.7 for valproate, 2.5 for carbamazepine, 1.5 for phenytoin and 1.2 for phenobarbitone.²⁸

---

1. Bloom BS. Daily regimen and compliance with treatment. BMJ. 2001 ↩

2. Nightingale CH, Murakawa T, Ambrose PC.  Antimicrobial pharmacodynamics in theory and clinical practice. Marcel Dekker Inc. New York. 2002 ↩

3. [1] El Moussaoui R, de Borgie CA, van den Broek P, et al. Effectiveness of discontinuing antibiotic treatment after three days versus eight days in mild to moderate-severe community acquired pneumonia: randomised double-blind study. BMJ 2007;332:1355-8 ↩

4. Costelloe C, Metcalfe C, Lovering A, et al. Effect of antibiotic prescribing in primary care on antimicrobial resistance in individual patients: systematic review and meta-analysis. BMJ2010;340:c2096. doi:10.1136/bmj.c2096 ↩

5. Butler CC, Simpson SA, Dunstan F, et al. Effectiveness of multifaceted educational programme to reduce antibiotic dispensing in primary care: practice based randomised controlled trial. BMJ 2012;344:d8173 doi: 10.1136/bmj.d8173 ↩

6. Raveh D, Kopyt M, Hite Y, et al. Risk factors for nephrotoxicity in elderly patients receiving once-daily aminoglycosides. Q J Med. 2002;95:291-7 ↩

7. Joynt GM, Lipman J, Gomersall CD, et al. The pharmacokinetics of once-daily dosing of ceftriaxone in critically ill patients. J Antimicrobial Chemotherapy 2001;47:421-9 ↩

8. McGee S, Abernethy WB, Simel DL. Is this patient hypovolaemic? JAMA. 1999;281:1022-1029 ↩

9. Brunkhorst FM, Engel C, Bloos F, et al. Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med. 2008;358:125-39 ↩

10. Evans L, Rhodes A, Alhazzani W, et al. Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock 2021. Crit Care Med. 2021 Nov 1;49(11):e1063-e1143. doi: 10.1097/CCM.0000000000005337. PMID: 34605781. ↩

11. Finfer S, Micallef S, Hammond N, et al, PLUS Study Investigators and the Australian New Zealand Intensive Care Society Clinical Trials Group. Balanced Multielectrolyte Solution versus Saline in Critically Ill Adults. N Engl J Med. 2022 Mar 3;386(9):815-826. doi: 10.1056/NEJMoa2114464. Epub 2022 Jan 18. PMID: 35041780. ↩

12. Wang P, Huang Y, Li J, Cao D, et al. Balanced crystalloid solutions versus normal saline in intensive care units: a systematic review and meta-analysis. Int Urol Nephrol. 2023 Apr 5. doi: 10.1007/s11255-023-03570-9. ↩

13. Hammond NE, Zampieri FG, Di Tanna GL, Garside T, Adigbli D, Cavalcanti AB, et al. Balanced crystalloids versus saline in critically ill adults – a systematic review with meta-analysis. NEJM Evid 2022;1 ↩

14. Cochrane Injuries Group Albumin Reviewers. Human albumin administration in critically ill patients: systematic review of randomised controlled trials. BMJ. 1998; 317: 235-240 ↩

15. The SAFE Study Investigators. A comparison of albumin and saline for fluid resuscitation in the intensive care unit. N Engl J Med. 2004; 350: 2247-2256 ↩

16. Doyle G, Furey S, Berlin R, et al. Gastrointestinal safety and tolerance of ibuprofen at maximum over-the-counter dose. Alimentary Pharm Therapeutics. 1999;13:897-906 ↩

17. Po AL, Zhang WY. Analgesic efficacy of ibuprofen alone and in combination with codeine in post-surgical pain: a meta-analysis. European  J Clin Pharm. 1998;53:303-11 ↩

18. Agra Y, Sacristan A, Gonzalez M, et al. J Pain and Symptom  Management 1998;15:1-7 ↩

19. Kredo T, Onia R. Pethidine – does familiarity or evidence perpetuate its use? S Afr Med J.2005;95:100-1 ↩

20. Kirkland SW, Cross E, Campbell S, Villa‐Roel C, Rowe BH. Intramuscular versus oral corticosteroids to reduce relapses following discharge from the emergency department for acute asthma. Cochrane Database of Systematic Reviews 2018, Issue 6. Art. No.: CD012629. DOI: 10.1002/14651858.CD012629.pub2  ↩

21. Lahn M, Bijur P, Gallagher EJ.  Randomized clinical trial of intramuscular vs. oral methylprednisolone in the treatment of asthma exacerbations following discharge from an emergency department. Chest.2004;126:362-8 ↩

22. CRASH trial collaborators. Effect of intravenous corticosteroids on death within 14 days in 10 008 adults with clinically significant head injury (MRC CRASH trial): randomised placebo-controlled trial. Lancet. 2004;364:1321-28. Update with 6 month outcomes: Lancet. 2005;365:1957-59 ↩

23. Sauerland S, Maegele M. A CRASH landing in severe head injury. Lancet. 2004;364:1291-2 ↩

24. Kopsaftis Z, Wood‐Baker R, Poole P. Influenza vaccine for chronic obstructive pulmonary disease (COPD). Cochrane Database of Systematic Reviews 2018, Issue 6. Art. No.: CD002733. DOI: 10.1002/14651858.CD002733.pub3 ↩

25. Demicheli V, Jefferson T, Ferroni E, Rivetti A, Di Pietrantonj C. Vaccines for preventing influenza in healthy adults. Cochrane Database of Systematic Reviews 2018, Issue 2. Art. No.: CD001269. DOI: 10.1002/14651858.CD001269.pub6 ↩

26. Moberley S, Holden J, Tatham DP, Andrews RM. Vaccines for preventing pneumococcal infection in adults. Cochrane Database of Systematic Reviews 2013, Issue 1. Art. No.: CD000422. DOI: 10.1002/14651858.CD000422.pub3 ↩

27. Purcell B, Samuelsson S, Hahne JM, et al. Effectiveness of antibiotics in preventing meningococcal disease after a case: systematic review. BMJ. 2004;328:1339-42 ↩

28. Kaufman DW, Shapiro S. Epidemiological assessment of drug-induced disease. Lancet. 2000;  356:1339-1341 ↩

Was this article helpful?

0 out Of 5 Stars

5 Stars		0%
4 Stars		0%
3 Stars		0%
2 Stars		0%
1 Stars		0%

5 Submit

Send Feedback

How can we improve this article?

Please submit the reason for your vote so that we can improve the article.

Submit