Comparison of national antimicrobial treatment guidelines, African Union

Abstract Objective To identify and compare antimicrobial treatment guidelines from African Union (AU) Member States. Methods We reviewed national government agency and public health institutes’ websites and communicated with country or regional focal points to identify existing treatment guidelines from AU Member States. We included guidelines if they contained disease-, syndrome- or pathogen-specific treatment recommendations and if those recommendations included antimicrobial name or class, dosage and therapy duration. The scope of the review was limited to infections and clinical syndromes that often have a bacterial cause. We assessed treatment guidelines for alignment with the Grading of Recommendations Assessment, Development and Evaluation (GRADE) criteria. We compared treatment recommendations for various common bacterial infections or clinical syndromes described across national guidelines and those described in three World Health Organization guidelines. Findings We identified 31 treatment guidelines from 20 of the 55 (36%) AU Member States; several countries had more than one treatment guideline that met our inclusion criteria. Fifteen (48%) guidelines from 10 countries have been published or updated since 2015. Methods used to develop the guidelines were not well described. No guidelines were developed according to the GRADE approach. Antimicrobial selection, dosage and duration of recommended therapies varied widely across guidelines for all infections and syndromes. Conclusion AU Member States lack antimicrobial treatment guidelines that meet internationally accepted methods and that draw from local evidence about disease burden and antimicrobial susceptibility.


Introduction
Antimicrobial resistance poses significant public health challenges and threatens the ability to treat many common infectious diseases. Across Africa, rising rates of drug resistance have been documented for the pathogens that cause tuberculosis, pneumonia, diarrhoeal diseases, malaria and sexually transmitted infections. 1 Documented levels of antimicrobial resistance rates are likely lower than the actual level due to limited surveillance activities and laboratory capacity and lack of consistent access and utilization of antimicrobial susceptibility testing. 2 Major drivers of antimicrobial resistance include: the misuse and overuse of antimicrobials in the human health and agricultural sectors; lack of access to clinically appropriate antimicrobials; lack of regulation and/or regulatory enforcement restricting access to antimicrobials to prescription-only use; and a higher burden of infectious diseases which is driven by low vaccination coverage and limited water, sanitation and hygiene infrastructure. [2][3][4] Given these risk factors for antimicrobial resistance, low-and middle-income countries are at a higher risk for the propagation of resistant pathogens compared to high-income countries. Moreover, the public health consequences of antimicrobial resistance may be higher in African countries because of failure to diagnose antimicrobial-resistant infections and the absence of expensive second-line therapies.
In 2018, the Africa Centres for Disease Control and Prevention (Africa CDC) released its Framework for Antimicrobial Resistance, a strategy to improve surveillance, delay emergence, limit transmission and mitigate harm from antimicrobial-resistant pathogens in Africa. 5 African Union (AU) Member States and stakeholders identified priority activities for implementing Africa CDC's framework. To delay the emergence and mitigate harm of antimicrobial resistance, many expert organizations recommended that antimicrobials only be used in accordance with established clinical guidelines that outline when, what and how to prescribe. However, representatives of the AU Member States noted that, except for select diseases such as human immunodeficiency virus (HIV) infection, tuberculosis and malaria, many health-care providers do not have country-specific guidelines and must rely on individual judgement or treatment guidelines developed outside of Africa.
Standardized treatment guidelines are an important tool to ensure the appropriate and optimal use of antimicrobials in the human health sector, particularly when informed by local data and combined with antimicrobial stewardship programmes. 6,7 We sought to identify existing standardized treatment guidelines from African government agencies or public health institutes, to assess their completeness and quality, and compare the antimicrobial treatment therapies described across the guidelines.

Methods
We reviewed the websites of health ministries, national public health institutes or equivalent national government agencies of all 55 AU Member States for relevant published standardized treatment guidelines. 8 We conducted two reviews, the first between December 2018 and March 2019 and the second between May and June 2021. For countries without functioning Research Antimicrobial treatment recommendations, Africa Jessica Craig et al.
health ministry websites or for countries where guidelines were not readily or apparently available, we contacted in-country Africa CDC focal points by email to help identify existing standardized treatment guidelines. In-country focal points were identified through Africa CDC Regional Collaborating Centres and included regional World Health Organization (WHO) staff. Two subsequent emails were sent to each focal person if immediate response did not occur. We were unable to identify specific in-country focal points in Djibouti, Libya, and Somalia; a regional focal point was contacted in these cases. In total, 15 national or regional focal points were contacted with two not responding.
To inform our work, we presented our proposed methods and inclusion criteria to a panel of 28 infectious disease clinicians and public health experts representing 13 AU Member States. This consultation resulted in exclusion of national and regional guidelines developed by professional bodies and organizations, given that these guidelines may not be widely or systematically available or used in health facilities across the continent. An evaluation of the access and utilization of guidelines developed by these bodies would be required to determine if they are consistently used by clinicians treating infectious diseases. Furthermore, these guidelines may not represent feasible clinical options at the national or continental level given challenges in access to various antimicrobials or other diagnostic or treatment paradigms.
We limited the scope of the review to bacterial infections and clinical syndromes that often have a bacterial cause. Disease-specific guidelines for HIV, malaria, tuberculosis and other infections or syndromes addressed by national or vertical disease control programmes in Africa were excluded. We considered guidelines published in any major continental language. For final review, we included guidelines only if they contained disease-, syndrome-or pathogen-specific treatment recommendations and recommendations included specific name or class of antimicrobial, dosage and duration of therapy.
Standardized treatment guidelines that met the inclusion criteria were assessed for alignment with the Grading of Recommendations Assessment, Development and Evaluation (GRADE) criteria. 9 This included a review of in-formation describing the methods used in guidelines development including whether the guidelines underwent external technical review or were informed by a systematic review of local, national or regional antimicrobial resistance or disease burden data. 9 We compiled antimicrobial treatment recommendations provided in each guideline for common bacterial diseases, including information on drug (or class of antimicrobial) selection, dosage and duration of therapy. Recommendations were segregated by adult, paediatric and/or neonatal patient populations as specified in the source guidelines. Special populations (e.g. pregnant women, patients with penicillin allergies) were included if noted in the source guidelines. Alternative and second-line therapy recommendations were also recorded. We did not extract other clinical information, such as case definition, recommended diagnostic testing and non-antimicrobial therapy treatments (e.g. vitamin supplements or pain management).
For comparison regarding first-and second-line antimicrobial selection, we also consulted the 2019 WHO Model list of essential medicines, 10 the 2019 WHO Model list of essential medicines for Children 11 and the Pocket book of hospital care for children, 2nd edition. 12 The WHO model lists only provided drug selection recommendations and did not provide recommended treatment dosages or durations; therefore, further comparison of recommended therapy was not possible. Finally, we compared treatment regimens provided across individual guidelines, including the WHO model lists, to assess variability in antimicrobial treatment recommendations. Table 1 presents the 31 standardized treatment guidelines from 20 countries that met the inclusion criteria.  Several countries had more than one published guideline included for final review: Ethiopia (2 guidelines), Kenya (2), the Seychelles (2), South Africa (5), Tunisia (3), Uganda (2) and Zambia (2). Despite extensive efforts, we did not identify existing standardized treatment guidelines for the remaining 35 AU Member States. Year of publication or last update ranged from 2001 to 2018; two standardized treatment guidelines did not state the year of publication. Only 15 (48%) of 31 standardized treatment guidelines from 10 countries have been published or updated since 2015. Four guidelines were published in French and the remaining were published in English. All guidelines were published by the health ministry or other national health agencies or research institutes. Twenty-three (74%) guidelines provided treatment recommendations for both adult and paediatric patient populations, five provided only adult-specific guidelines and three provided paediatric-only recommendations. Overall, the description methods used in the development of the guidelines was poor and no guidelines were developed according to the GRADE approach. Only three (10%) guidelines reported that the treatment recommendations considered available antimicrobial resistance data, while 10 (32%) guidelines stated that data on local disease burden was considered when developing recommendations.

Description of guidelines
We extracted and compiled treatment regimen recommendations for 27 infections and/or syndromes for adult patients and 26 for paediatric patients (Table 2). Meningitis, conjunctivitis, mild, moderate or non-severe pneumonia and urinary tract infection were the most common bacterial infections or clinical syndromes included in guidelines for adult patients; each were covered by 14 (70%) countries. Thirteen (65%) countries had guidelines that provided treatment recommendations for adult patients for acute otitis media, cholera, impetigo and pelvic inflammatory disease, while guidelines from 12 (60%) countries provided recommendations for acute and/or chronic bronchitis, syphilis and typhoid or enteric fever. Cholera and meningitis were the most frequently included infections or clinical syndromes for paediatric patients followed by acute otitis media, conjunctivitis, impetigo and typhoid or enteric fever. The WHO model list provided treatment regimen recommendations for adult populations for 48% (13/27) of the bacterial infections or syndromes commonly included in the standardized treatment guidelines of AU Member States. The WHO model list for children and the pocket book provided recommendations for 50% (13) and 27% (7), respectively, of the 26 infections or syndromes covered by the guidelines.

Research
Antimicrobial treatment recommendations, Africa Jessica Craig et al.

Comparison of regimens
Antimicrobial selection and dosage and duration of recommended therapies varied across guidelines for all bacterial infections or clinical syndromes included in this analysis. Few guidelines provided organism-specific treatment recommendations or guidance on tailoring antimicrobial therapy according to bacterial culture or antimicrobial susceptibility testing results. Complete data sets for all infections and clinical syndromes are available in the authors' data repository. 44 For brevity, we present illustrative examples below showing the range of recommendations: a comparison of treatment recommendations for bacterial meningitis in paediatric and neonatal patient populations and for bronchitis in adult patients.

Paediatric population
A total of 14 countries provided recommendations for the treatment of acute bacterial or suspected bacterial meningitis in paediatric and/or neonatal patient populations (Table 3). Of those, eight (57%) countries provided organism-specific drug selection recommendations or provided organismspecific therapy durations whereas the Antimicrobial treatment recommendations, Africa Jessica Craig et al.
other standardized treatment guidelines did not specify a causative agent or provided recommendations in situations when the causative agent was unknown. The WHO model list for children also provided treatment recommendations for acute bacterial meningitis with no organism-specific recommendations where the first-choice drug selections were cefotaxime for neonates and ceftriaxone for children. Second-choice antimicrobials included meropenem for neonates and amoxicillin, ampicillin, benzylpenicillin and chloramphenicol for children older than 2 years. Across the guidelines that provided treatment recommendations for bacterial meningitis in paediatric patients (excluding neonates) where the causative agent was not specified or in cases where it is unknown, recommended drug selections for monotherapy included benzylpenicillin, ampicillin, ceftriaxone and chloramphenicol. The recommended combination therapies were benzylpenicillin plus chloramphenicol. For neonates, recommended drug selections for monotherapy included ceftriaxone and fluconazole, and combination therapies included ampicillin IV plus gentamicin and benzylpenicillin plus chloramphenicol. Multiple guidelines recommended the same first-choice antimicrobial; however, dosage and therapy duration varied. For instance, dosage and duration recommendations for monotherapy treatment of paediatric patients (excluding neonates) with ceftriaxone ranged from 100 mg/kg daily for 7 days to 100 mg/kg daily for 10 to 14 days and 50-100 mg/kg every 12 hours for 14 days.
The variation between guidelines was less when the causative microorganism of the infection or syndrome was specified. For example, five guidelines provided recommendations for the treatment of acute meningitis in paediatric patients (excluding neonates) caused by Streptococcus pneumoniae. All five guidelines recommended monotherapy with either benzylpenicillin or ceftriaxone for 10 to 14 days; dosage recommendations were similar across the four guidelines that provided such information (Table 3).
Of the 14 countries that had national guidelines for the treatment of meningitis, only six (43%) included guidance or instructions for obtaining cultures and antimicrobial susceptibility testing or tailoring drug selection or duration accordingly (available in data repository). 44 For example, the Standard treatment guidelines and essential medicines list for South Africa, hospital level paediatrics provide recommendations for empirical treatment of meningitis and instruct users to "Adjust antimicrobial therapy according to culture and sensitivity" and to "Re-assess antimicrobial therapy when blood and CSF [cerebrospinal fluid] culture and sensitivity results become available, or when improvement is not evident within 72-96 hours." 29 The Gambia standard drug treatment guide provides a recommendation to treat paediatric patients older than 2 months of age and presenting with meningitis with a combination therapy of benzylpenicillin plus chloramphenicol. 16 Alongside the antimicrobial treatment recommendation, there is also guidance to "Continue until culture result is known, after which use a single drug therapy with chloramphenicol, benzylpenicillin or ampicillin." 16

Adult population
There were discrepancies among guidelines regarding the use of antimicrobials for the treatment of certain conditions for adult population. For example, guidelines from eight countries described recommendations for the treatment of acute bronchitis. Of those, three recommended treatments with various antimicrobial therapy regimens including 250-500 mg of amoxicillin for 5 days, 500 mg of amoxicillin every 8 hours for 7 days, 200 mg of doxycycline on the first day of treatment followed by 100 mg once a day for 5 days, and 500 mg of amoxicillin/clavulanic acid every 12 hours for 7 days (Table 4). Four other guidelines stated that antimicrobial use was not indicated for the treatment of acute bronchitis while a fifth guideline stated that antibiotics were not indicated for uncomplicated bronchitis but for purulent bronchitis, treatment with amoxicillin at a dosage of 500 mg every 8 hours for 4 or more days was indicated. All four guidelines for the treatment of chronic bronchitis indicated treatment with an antimicrobial agent. The WHO model list 10 did not provide clinical guidance for the treatment of acute or chronic bronchitis in adults.

Discussion
The misuse and overuse of antimicrobials in the human health sector is a major driver of antimicrobial resistance globally. Per capita antimicrobial consumption in Africa and other low-and middleincome countries is rising at a faster rate than high-income countries. 45 In part, this consumption pattern may be due to a lack of clear guidance for the use of antimicrobials in the clinical setting, improved access to antimicrobials, and non-prescription use of antimicrobials. 45 Here, we found that only one third of Only about half of the bacterial infections and/or syndromes covered in existing standardized treatment guidelines from AU Member States across adult and paediatric patient populations were addressed in WHO model lists 10,11 and less than a third were included in the Pocket book of hospital care for children. 12 This finding indicates a potential discrepancy between disease burden and priorities at the international and regional contexts and reinforces the need for context-specific standardized treatment guidelines to guide the appropriate clinical treatment of these infections and clinical syndromes. Finally, the recommended drug selection, dosage and duration of therapy varied across the guidelines, indicating a lack of clear, consensus clinical guidance and potential for the misuse of antimicrobials.
Clinical guidelines that provide explicit recommendations for the treatment of infectious diseases and aid in clinical decision-making have been shown to reduce inappropriate antimicrobial prescribing as well as improve the quality of care. 46,47 High quality treatment guidelines should be based on a rigorous multidisciplinary evaluation of all available scientific evidence about treatment effectiveness and local disease and antimicrobial resistance burden. 7,48,49 Moreover, as antimicrobial resistance increasingly poses a public health threat, guidelines must also encourage clinicians to obtain cultures, identify the causative microorganism and conduct antimicrobial susceptibility testing before or during therapeutic intervention.
This study had several limitations. First, we did not include guidelines developed by nongovernmental public or private institutions, professional organizations or individual health facilities in this review, which may have limited the number of total countries represented in this study. Moreover, guidelines developed outside of the national government infrastructure may represent an important source for information guiding clinical decision-making and antimicrobial prescribing practices at the facility, national and regional levels. While we made every effort to identify all existing guidelines for each eligible country, some guidelines used in clinical practice may not be publicly available or easily searchable online and may have been omitted from inclusion. In the future, all efforts should be exhausted to ensure that all applicable guidelines are identified and considered for inclusion; maintenance of a guidelines database may also be useful. In addition, this study does not account for the awareness of, utilization of and/or adherence to standardized treatment guidelines at the local or national levels nor other factors that may influence feasibility of using guidelines, such as availability and cost of recommended antimicrobials. As such, antimicrobial treatment regimens Co-trimoxazole (960 mg every 12 hours) or amoxicillin (500 mg over 8 hours)

days
Zimbabwe 43 Treat with antibiotics (amoxicillin 500 mg three times/day) or doxycycline (100 mg once/day for 7 days) if sputum colour has changed to purulent, if there is fever or new chest X-ray infiltrates 7 days a We summarized treatment recommendations from the standard treatment guidelines and edited for grammar, clarity and/or length.

Research
Antimicrobial treatment recommendations, Africa Jessica Craig et al.
employed in clinical practice may vary widely from those recommended in any standardized treatment guidelines. In addition to the suboptimal treatment guidelines, the high antimicrobial consumption rates in the African Region may reflect the higher burden of infectious diseases, the high prevalence of substandard pharmaceutical agents, the lack of access to health-care services and lack of regulation around prescriptiononly use of antimicrobials. These factors lead to high rates of self-medication and non-prescription antimicrobials sales and consumption. 50 Therefore, while the development and implementation of standardized treatment guidelines offer an opportunity to reduce the misuse and overuse of antimicrobials in the human health sector, other antimicrobial stewardship interventions that address other drivers of antimicrobial overuse and misuse and the emergence and spread of antimicrobial resistance more broadly must also be implemented.
In conclusion, AU Member States need to develop standardized treatment guidelines that meet internationally accepted standards for methods and are based on locally derived clinical evidence, disease burden and resistance profiles. To achieve this, countries must also increase their capacity for evidence building, including antimicrobial resistance and infectious disease surveillance. Furthermore, many countries across the continent lack adequate access to clean water and sanitation facilities, and have vaccination coverage below recommended levels. In addition to ensuring antimicrobial treatment is appropriate, preventing infections where possible is essential to reduce antimicrobial use, slow the emergence of resistance and ultimately improve health outcomes and save lives. ■

Research
Antimicrobial treatment recommendations, Africa Jessica Craig et al.
Conclusión Los Estados miembros de la UA carecen de directrices de tratamiento antibiótico que cumplan con los métodos aceptados a nivel internacional y que se basen en las evidencias locales sobre la carga de las enfermedades y la sensibilidad a los antibióticos.