Overdiagnosis and overprescription of infectious diseases in primary

Overdiagnosis and overprescription of infectious diseases in primary care Carles Llor, GEIAP-SEIMC Via Roma Health Centre, Barcelona Københavns Universitet, Copenhagen Salamanca, 25 October 2019

Competing interests Research support Employee, consultant, stakeholders, speakers bureau, honoraria I have received research grants from Abbott Diagnostics No other support from private companies None

3 Learning objectives • Recognise the huge problem of the antimicrobial resistance • Recognise the huge amount of antibiotic overprescribing in primary care • Identify myths most clinicians believe which results in inappropriate antibiotic prescribing • Describe strategies targeting clinicians to better prescribe antibiotics • Wrapping up

Effect of prior antimicrobial courses on the acquisition of resistance Costelloe C et al. BMJ 2010; 340: c 2096.

Consumption of antibiotics in 76 countries in 2015 Klein EY et al. Proc Natl Acad Sci U S A. 2018; 115: E 3463– 70.

Sir Luke Fildes’s painting ‘The Doctor’, 1891

Defining ideal antibiotic prescribing Smith DRM et al. J Antimicrob Chemother 2018; 73(Suppl 2): ii 11– 8. 8

05/06/2021 9 Expected number of events over 10 years in a hypothetical high antibiotic prescribing general practice with 7, 000 patients Gulliford MC et al. BMJ 2016; 354: i 3410.

Myths in common infectious diseases in primary care Recommend broad-spectrum antibiotics in bacterial infections Recommend standard antibiotic courses instead of shorter regimens for respiratory tract infections The lower airways and the urinary tract are sterile under normal conditions Request radiological study for the diagnosis of acute rhinosinusitis The diagnosis of group A β-haemolytic streptococcus infection is based on symptoms and signs Pharyngeal exudate is synonymous of streptococcal aetiology Prescribe antibiotics for patients with negative rapid antigen detection test results Acute otitis media should be routinely treated with antibiotics Prescribe antibiotics in acute cough and purulent sputum in patients without COPD The diagnosis of pneumonia is based on symptoms and signs

Myths in common infectious diseases in primary care Recommend dual antibiotic treatment in community-acquired pneumonia Treat asymptomatic bacteriuria in patients with diabetes mellitus Perform dipsticks in patients with indwelling catheters or institutionalised patients without symptoms of UTI Administer prophylactic antibiotic therapy when an indwelling catheter must be replaced Perform a dipstick test in a woman with clear symptoms of acute cystitis Treat with antibiotics when the urine is cloudy and smells bad The presence of nitrites in the urine is predictive of UTI Intensive antibiotic treatment is required when Pseudomonas aeruginosa is isolated in an ear swab Systemic treatment is more effective than topical eardrops in otitis externa Dental phlegmons should be treated with antibiotics

Asymptomatic bacteriuria, a prevalent condition Population group % bacteriuria Healthy premenopausal women 1. 0 – 5. 0 Pregnancy Postmenopausal women (50 -70 years) 1. 9 – 9. 5 2. 8 – 8. 6 Diabetic patients: - Men - Women 0. 7 – 19 9 – 27 Community-dwelling elderly: - Men - Women 3. 6 – 19 10. 8 – 16 Institutionalization: - Men - Women 15 – 40 25 – 50 Spinal cord lesions with intermittent indwelling urinary cathether use 23 – 89 Hemodialysis Indwelling urinary catheter: - Short-term - Long-term 28 9 – 23 100

Antibiotic therapy in asymptomatic bacteriuria Group A: not treated; Group B: treated T et al. Clin Infect Dis 2012; 55: 771– 7. ; T et al. Clin Infect Dis 2015; 61: 1655– 61. 1 Cai 2 Cai

Role of the urine characteristics Characteristics of the urine • Change in the colour of the urine (darker) • Change in the urine smell (stinking) • Change in the urine turbidity (cloudy) Turbidity of urine for the diagnosis of UTI: Se 13. 3%, Sp 96. 5%, PPV 40%, NPV 86. 3% The odour depends on the hydration and urea concentration in urine Avoid using dipsticks if the presence or absence of symptoms is clear Foley A et al. J Am Board Fam Med 2011; 24: 474– 5. ; Nicolle LE. Infect Control Hosp Epidemiol 2001; 22: 316– 21. ; Lammers RL et al. Jann Emerg Med 2001; 38: 505– 12. ; Schulz L et al. J Emerg Med 2016; 51: 25– 30.

Leukocyte-esterase and nitrites in dipsticks Leukocyte-esterase • Its presence is not diagnostic of UTI • PPV of UTI: 18 – 75% • Present in 90% of asymptomatic bacteriuria • A positive leukocyte-esterase can be elevated in elderly, non-infectious hematuria, acute kidney failure, sexual transmitted infections, non-infectious cystitis • Its absence does not rule out an UTI Leukocyte-esterase and nitrites Nitrites • Most predictive for UTI • Not all the uropathogens result in positive nitrites: Pseudomonas spp. , Enterococcus spp. • Both positive: Se 48% i Sp 93% of UTI in institusionalised • Both negative and negative blood: NPV 73% Juthani-Mehta Et al. Infect Control Hosp Epidemiol 2007; 28: 889– 91. ; Nicolle NE. Infect Control Hosp Epidemiol 2001; 22: 167– 75. ; Schulz L et al. J Emerg Med 2016; 51: 25– 30. ; Deville WL et al. BMC Urol 2004; 4: 4. ; Schulz L et al. J Emerg Med 2016; 51: 25– 30. ; Little P et al. Br J Gen Pract 2006; 56: 606– 12.

Chest X-rays for control of pneumonia: Recovery in pneumonia 100 90 80 70 60 % 50 40 30 20 10 0 Day 10 Clinical cure Clinical score normalised Day 28 Radiological resolution Bruns AH et al. J Gen Intern Med 2009; 25: 1182 -7.

Sinus X-ray for the diagnosis of bacterial rhinosinusitis Prediction of signs and symptoms for the diagnosis of bacterial rhinosinusitis Signs and symptoms Sensit. Specif. Pos. predict. value Neg. predict. value Purulent rinorrhoea 35% 78% 62% 78% Pain when bending the head 75% 77% 78% 73% Maxillar tooth pain 66% 49% 56% Symptoms after a RTI 89% 79% 83% 87% Nasal obstruction 60% 22% 53% 15% Diagnosis of rhinosinusitis Test Sensit. Specif. Simple X-ray Variable Ultrasound High Variable CT High Low MRI High Low Punction High Clinical High Moderate assessment Piccinillo JF et al. N Engl J Med 2004; 351: 902 -10.

Accuracy for history and physical examination elements in the diagnosis of GABHS sore throat Symptoms and signs Sensitivity Specificity Positive likelihood ratio Negative likelihood ratio Any exudates 0. 21 -0. 58 0. 69 -0. 92 1. 5 -2. 6 0. 66 -0. 94 History of fever 0. 3 -0. 92 0. 23 -0. 90 0. 97 -2. 6 0. 32 -1. 0 Temperature >37. 8ºC 0. 11 -0. 84 0. 43 -0. 96 1. 1 -3. 0 0. 27 -0. 94 Cervical nodes 0. 55 -0. 82 0. 34 -0. 73 0. 47 -2. 9 0. 58 -0. 92 Tonsillar inflamation 0. 56 -0. 86 1. 4 -3. 1 0. 56 -0. 72 Tender cervical nodes 0. 32 -0. 66 0. 53 -0. 84 1. 2 -1. 9 0. 49 -0. 71 No cough 0. 51 -0. 79 0. 36 -0. 68 1. 1 -1. 7 0. 53 -0. 89 Duration < 3 days 0. 26 -0. 93 0. 54 -0. 64 0. 72 -3. 5 0. 15 -2. 2 Ebell MH et al. JAMA 2000; 284: 2912– 8.

Clinical scales for the diagnosis of group A streptococcus strep throat Criteria Centor 1 Temp. >38. 5ºC in the last 24 hours Fever. PAIN 3 +1 +1 +1 Temp. >38ºC in the last 24 hours Tonsillar exudate Mc. Isaac 2 +1 Tonsillar exudate or inflammation +1 +1 Inspection with pus +1 Tender cervical nodes +1 +1 No cough +1 +1 Age 3 -14 years +1 Age >44 years -1 Rapid visit to the GP (≤ 3 days) 1 Mc. Isaac +1 Number of Probability Mc. Isaac of GABHS criteria infection 4 -5 39 -57% 3 25 -35% 2 10 -17% 1 10% 0 2. 5% +1 WJ et al. CMAJ 2000; 163: 811– 5. 2 Centor RM et al. Med Decis Making 1981; 1: 239– 46. 3 Little P et al. BMJ 2013; 347: f 5806.

20 Significant differences in the diagnosis of pneumonia in two different European countries Christensen SF et al. Prim Care Respir J. 2013; 22: 454– 8.

Acute rhinosinusitis Acute viral rhinosinusitis (common cold) Acute bacterial rhinosinusitis Post-viral rhinosinusitis Fokkens W et al. Rhinology 2012; 50(Suppl 23): 1– 198.

Management of acute rhinosinusitis Comparison of different treatments • Double blind placebo-controlled randomised clinical trial with 240 adults • Symptoms included: unilateral or bilateral nasal purulent secretion or unilateral local pain • Management: antibiotic or/and steroid therapy • Outcome: symptoms on day 10 Managed with antibiotics 29% Managed with placebo antibiotics 34% Managed with nasal steroids 31% Managed with placebo nasal steroids 31% Williamson IG et al. JAMA 2007; 298: 2487– 96.

Benefit of antibiotics for LRTI Little P et al. Lancet Infect Dis 2013; 13: 123– 9.

Benefit of antibiotics for dental infections • Antibiotic therapy rarely indicated in cavities, pulpitis, periodontal disease, gingivitis • Antibiotic recommended in immunocomprimsed patients and in case of fever and intraoral tumefaction and in pericoronitis • Good hygiene and correct brushing is crucial • Recommend fluorides (for prevention and progression) in toothpaste • Reducing simple sugars Robles Raya P et al. Aten Primaria 2017; 49: 611 -8.

Short antibiotic regimens for pneumonia n Age Antibiotics RR clinical success 5 vs. 10 101 - Azithromycin vs. Erythromycin 1. 16 (0. 63 – 2. 14) Kinasewitz G, 19912 5 vs. 10 119 42 yr. Azithromycin vs. Cefaclor 1. 06 (0. 70 - 1. 62) Schonwald S, 19943 3 vs. 10 150 49 yr. Azithromycin vs. Roxithromycin 0. 13 (0. 03 – 0. 59) Kobayashi H, 19954 3 vs. 14 163 - Azithromycin vs. Clarithromycin 0. 93 (0. 58 -1. 50) O’Doherty B, 19985 3 vs. 10 203 51 yr. Azithromycin vs. Clarithromycin 1. 01 (0. 56 – 1. 83) Leophonte P, 20026 5 vs. 10 244 64 yr. Ceftriaxone 0. 90 (0. 59 – 1. 35) Dunbar LM, 20037 5 vs. 10 528 54 yr. Levofloxacin 0. 80 (0. 62 – 1. 03) Leophonte P, 20048 7 vs. 10 320 54 yr. Gemifloxacin vs. Amox/clav. 1. 09 (0. 72 – 1. 65) Rahav G, 20049 3 vs. 10 108 50 yr. Azithromycin vs. Various 0. 12 (0. 02 – 0. 99) Tellier G, 200410 5 vs. 10 559 42 yr. Azithromycin vs. Clarithromycin 0. 94 (0. 65 – 1. 37) 2, 796 40 -64 yr. Author, year Days Schonwald S, 19901 TOTAL 0. 89 (0. 46 – 1. 43) S et al. J Antimicrob Chemother 1990; 25(Suppl A): 123– 6. 2 Kinasewitz G et al. Eur J Clin Microbiol Infect Dis 1991; 10: 872– 7. 3 Schonwald S et al. Scand J Infect Dis 1994; 26: 706– 10. 4 Kobayashi H et al. Jpn J Chemother 1995; 43: 757– 74. . 5 O’Doherty B et al. J Clin Microbiol Infect Dis 1998; 17: 822– 33. . 6 Leophonte P et al. Med Mal Infect 2002; 32: 369– 81. 7 Dunbar LM et al. Clin Infect Dis 2003; 37: 752– 60. 8 Leophonte P et al. Respir Med 2004; 98: 708– 20. 9 Rahav G et al. Int J Antimicrob Agents 2004; 24: 181– 4. 10 Tellier G et al. J Antimicrob Chemother 2004; 54: 515– 23. 1 Schonwald

Short antibiotic regimens for pneumonia: Clinical failure Li JZ et al. Am J Med 2007; 120: 783– 90.

28 Importance of the clinical stability criteria Body temperature ≤ 37. 8 for 48 hours Clinical stability: - Blood pressure > 90/60 mm Hg Respiratory rate < 24 bpm Heart rate < 100 bpm Arterial oxygen saturation > 90%

Short antibiotic regimens for acute exacerbations of COPD • Adults � 18 yr. • Clinical diagnosis of COPD or chronic bronchitis exacerbation • Not treated with antibiotics at the moment of the diagnosis • Randomised allocation to an antibiotic therapy � 5 dies vs. treatment >5 days • Double blind clinical trial El Moussaoui R et al. Thorax 2008; 63: 415– 22.

Short antibiotic regimens for rhinosinusitis Falagas ME et al. Br J Clin Pharmacol 2008; 67: 161– 71.

31 Effectiveness of shorter therapies for sore throat

Short courses that should be recommended Infection Comparison Acute bacterial rhinosinusitis 1 3 -7 vs. 6 -10 days Number of Evidence patients 4, 430 OR of clinical cure: 0. 95 (0. 81 – 1. 12). A 5 -day (12 studies) regimen present a clinical success rate similar to 10 -day courses 5 Acute otitis media 2 570 (5 studies) 1, 540 (8 studies) OR of therapy failure of 0. 85 (0. 60 - 1. 21) in children older than 2 years old RR of clinical failure: 0. 96 (0. 74 – 1. 26). In another study, the clinical success was similar between 3 day and 8 -day regimens <5 vs. ≥ 5 days 10, 698 (21 studies) OR of clinical cure at day 25: 0. 99 (0. 90 – 1. 08) 7 -14 vs. 14 -42 7 days 185 (2 studies) OR of clinical success: 1. 03 (0. 80 – 1. 32) 5 Community-acquired pneumonia 3, 4 2 -7 vs. ≥ 7 days 3 -5 vs. >7 days 5 COPD acute exacerbation 5 Acute pyelonefritis 6 5 ME et al. Br J Clin Pharmacol 2009; 67: 161– 71. 2 Kozyrskyj A et al. Cochrane Database Syst Rev 2010; 9: CD 001095. 3 Li JZ et al. Am J Med 2007; 120: 783– 90. 4 el Moussaoui R et al. BMJ 2006; 332: 1355. 5 El Moussaoui R et al. Thorax 2008; 63: 415– 22. 6 Kyriakidou KG et al. Clin Ther 2008; 30: 1859– 68. 1 Falagas

Critically important antibiotics

Interventions aimed at reducing unnecessary antibiotic prescribing targeting healthcare professionals Intervention Level of evidence Quality indicators Education Low Limiting the over-the-counter sale Communication skills training + leaflets Medium Education material for patients Low Guidelines Low Clinical decision support systems Low Delayed antibiotic prescribing Medium Pay-for-performance Public commitment Low Sickness leave regulation Point of care tests Medium Audit and feedback Low Unit dispensing of antibiotics Reducing the number of available antibiotics Increasing the price of antibiotics Separating prescribing and dispensing Prescribers’ remuneration system Limiting advertising of antibiotics ? Dyar OJ et al. Clin Microbiol Infect 2017; 23: 793– 8. Restrictive prescribing measures Low

Prediction of the risk of pneumonia Patients with lower respiratory tract infections (n=3, 106) Chest X-rays not performed (n=258) Patients with chest radiographs available (n=2, 848) Chest radiographs of insufficient quality (n=28) Patients with chest radiographs of sufficient quality (n=2, 820) Signs and symptoms: - Absence of runny nose - Dyspnoea - Crackles - vesicular breathing - Tachycardia (> 100 bpm) - Temperature >37. 8 C Van Vugt SF et al. BMJ 2013; 346: f 2450.

Incidence of pneumonia depending on CRP concentrations (mg/l) 34/96 24/159 55/2, 249 11/227 16/246 Van Vugt SF et al. BMJ 2013; 346: f 2450.

Association between use of rapid tests and antibiotic consumption Consumption Rapid tests Consumption No tests

Recommended management for LRTIs Anamnesis and clinical examination: LRTI Uncomplicated RTI Additional research not required Antibiotics not indicated <20 mg/L ( 70%) Complicated RTI (suspected pneumonia) Moderately sick patient Seriously ill patient CRP -POCT Rule out pneumonia 20 -100 mg/L ( 25%) >100 mg/L ( 5%) Consider age, comorbidity, risk factors. Consider delayed antibiotic prescribing in some cases

Take-home messages about rapid tests 1. If you decide to treat a patient with an acute pharyngitis with antibiotics, then you are requested to perform a rapid antigen detection test for group A β-haemolytic streptococcus You will realise that most tests are negative. . . 2. If you choose to treat antibiotics a patient with a lower respiratory tract without a radiological confirmation of pneumonia, then you are requested to perform a Creactive protein rapid test You will realise that most determinations give results below 20 mg / l. . .

To test or not to test Added value of point-of-care tests increases Perceived patient demand weak Clinical uncertainty Clinical certainty Perceived patient demand strong

Delayed antibiotic prescribing Meta-analysis of 11 studies, with a total of 3, 555 participants. Outcome Immediate Delayed prescribing Shorter duration of symptoms ● ● Lower presence of complications ● ● Reduction in antibiotic utilisation ● Lower patient reattendance ● Greater patient satisfaction ● No antibiotics ● ● Spurling GKP et al. Cochrane Database Syyst Rev 2017; 9: CD 004417.

Example Recomendaciones para el tratamiento de su infección Nombre del paciente: ……………………………………………………. Nombre del médico: ……………………………………………………. La infección que tiene Ud. (o su hijo) Dura normalmente Infección de oído medio 4 días Dolor de garganta 7 días Resfriado o catarro 10 días Sinusitis 18 días Tos o bronquitis 3 semanas Otra infección: ……………… …… días Consejos que puede seguir para tratar mejor la infección Cuándo debería buscar ayuda? Contacte con su médico o llame a urgencias • Descanse. • Beba líquidos. • Pregunte a su farmacéutico por medicamentos que puedan aliviar los síntomas. • La fiebre es un signo de que su cuerpo está luchando contra la infección y se resuelve espontáneamente en la mayor parte de casos. Puede usar paracetamol (o ibuprofeno). • Lávese bien las manos para no diseminar la infección. …………………… Los siguientes puntos representan signos de enfermedad grave que obligan a ser valorados de forma urgente: • Si presenta un dolor de cabeza muy intenso. • Si su piel está muy fría y tiene un color extraño, o se presenta una erupción rara. • Si presenta confusión. • Si presenta dificultad para respirar. • Si presenta dolor fuerte en el pecho. • Si presenta dificultades para la deglución. • Si expectora sangre. • Si empeora su enfermedad. • ……………………………………. . Debe recoger la prescripción antibiótica después de días si Ud. (o su hijo) no se encuentra mejor o se encuentra peor. Retirar la receta en : La recepción del centro La consulta de su médico Otras indicaciones (p. ej en caso de receta electrónica: ) • • • La mayor parte de resfriados, episodios de tos, infecciones de oído, dolor de garganta y otras infecciones mejoran generalmente sin tratamiento antibiótico, ya que su cuerpo lucha generalmente contra estas infecciones de forma natural. Si toma antibióticos cuando no los necesita hace que las bacterias se vuelvan resistentes. Esto significa que es más probable que estos fármacos no sean tan efectivos en un futuro, cuando Ud. realmente los necesite tomar. Los antibióticos pueden provocar efectos secundarios, como erupción cutánea, vaginitis, dolor de estomago, diarrea, reacciones a la luz del sol y otros síntomas, o que enferme por mezclar algunos antibióticos con bebidas alcohólicas. No comparta nunca los antibióticos y devuelva las dosis de antibiótico que no ha usado a la farmacia

Communication skills enhancement + leaflets Setting realistic expectations for symptom duration Condition Average duration of symptoms Acute otitis media 4 days Acute sore throat 1 week Common cold 1½ weeks Acute rhinosinusitis 2½ weeks Acute cough/bronchitis 3 weeks

Antibiotic overprescribing in respiratory tract infections Only 20% of the antibiotics we give are really indicated! 44

Antibiotics: “one of the most uncomfortable prescribing decisions general practitioners make” Bradley CP. BMJ 1992; 304: 294– 6. 45

Conclusions 1. GPs erroneously believe that some myths are still true and this results in a huge amount of inappropriate antibiotic prescribing behaviour. 2. The ideal antibiotic prescribing rate should be revised on the basis of the latest medical evidence published. 3. We don’t treat results (urine cultures), we treat patients. 4. Shorter antibacterial therapies are preferable in many urinary and respiratory tract infections as they have been proved to be as effective as longer courses. 5. Use of some rapid tests in uncertain conditions, delayed antibiotic prescribing and better communication skills help clinicians optimize antibiotic prescribing.

Questions and suggestions welcome! carles. llor@gmail. com