ABSTRACT

Objectives.

To determine the level of adherence to clinical guidelines in prescribing amoxicillin to children younger than 5 years with pneumonia in outpatient settings in Colombia from 2017 to 2019, and assess the factors associated with adherence

Methods.

This was a cross-sectional study of secondary data from the Colombian Integrated Social Protection Information System database. Adherence was defined as prescription of oral amoxicillin for bacterial and unspecified pneumonia and non-prescription for viral pneumonia. Variables examined included: age (< 1 year, 1–4 years) of child; sex; cause of pneumonia (bacterial, viral, unspecified); region (Andean, Amazonian, Pacific, Caribbean, Insular, Orinoquian); and payment mechanism (without prior authorization, capitation, direct payment, pay per case, pay for event).

Results.

Of 215 925 cases of community-acquired pneumonia reported during 2017–2019, 64.8% were from the Andean region, 73.9% were bacterial pneumonia and 1.8% were viral pneumonia. Adherence to guidelines was observed in 5.8% of cases: this was highest for children diagnosed with viral (86.0%) compared with bacterial (2.0%) pneumonia. For children diagnosed with bacterial pneumonia, 9.4% were prescribed any antibiotic. A greater proportion of children covered by capitated payments (22.3%) were given treatment consistent with the guidelines compared with payment for event (1.3%).

Conclusion.

In this first study from Colombia, adherence to guidelines for outpatient treatment of children with bacterial pneumonia was low and was better for viral pneumonia. Further qualitative studies are needed to explore the reasons for this lack of adherence and why bacterial pneumonia was the most commonly reported etiology.

Keywords
Pneumonia; child; outpatients; amoxicillin; guideline adherence; Colombia

RESUMEN

Objetivos.

Determinar el nivel de adherencia a las directrices clínicas al momento de prescribir amoxicilina a menores de 5 años con neumonía en entornos de atención ambulatoria en Colombia entre el 2017 y el 2019, así como evaluar los factores asociados con la adherencia.

Métodos.

Este fue un estudio transversal de datos secundarios de la base de datos del Sistema Integral de Información de la Protección Social de Colombia. La adherencia se definió como la prescripción de amoxicilina por vía oral para las neumonías bacterianas y no especificadas, y la ausencia de prescripción para las neumonías virales. Las variables examinadas incluyeron: edad (< 1 año, 1 a 4 años); sexo; causa de la neumonía (bacteriana, viral, no especificada); región (andina, amazónica, Pacífico, Caribe, insular, Orinoco); y mecanismo de pago (sin autorización previa, capitación, pago directo, pago por caso, pago por evento).

Resultados.

De 215 925 casos de neumonía adquirida en la comunidad notificados durante el período 2017-2019, el 64,8% correspondieron a la región andina, el 73,9% a neumonía bacteriana y el 1,8% a neumonía viral. Se observó la adherencia a las directrices en el 5,8% de los casos: esta cifra fue más alta para la población infantil diagnosticada con neumonía viral (86,0%) que para la diagnosticada con neumonía bacteriana (2,0%). En el caso de la población infantil diagnosticada con neumonía bacteriana, al 9,4% se le recetó algún antibiótico. La proporción de población infantil cubierta por pagos capitados (22,3%) que recibió un tratamiento en consonancia con las directrices fue mayor que la de la población cubierta por pagos por evento (1,3%).

Conclusión.

En este primer estudio de Colombia, la adherencia a las directrices sobre el tratamiento ambulatorio de la población infantil con neumonía bacteriana fue bajo, en tanto que resultó superior en el caso de la neumonía viral. Se necesitan más estudios cualitativos para indagar sobre los motivos de esta falta de adherencia y las razones por las cuales la neumonía bacteriana fue la etiología notificada con mayor frecuencia.

Palabras clave
Neumonía; niño; pacientes ambulatorios; amoxicilina; adhesión a directriz; Colombia

RESUMO

Objetivos.

Determinar o nível de adesão às diretrizes clínicas para prescrição de amoxicilina em regime ambulatorial para crianças menores de 5 anos com pneumonia na Colômbia, de 2017 a 2019, e avaliar os fatores associados à adesão.

Métodos.

Estudo transversal de dados secundários do banco de dados do Sistema Integrado de Informação sobre Proteção Social da Colômbia. Definiu-se adesão como prescrição de amoxicilina oral para pneumonia bacteriana e não especificada, e não prescrição para pneumonia viral. As variáveis examinadas incluíram: idade da criança (< 1 ano, 1–4 anos), sexo, etiologia da pneumonia (bacteriana, viral, não especificada), região (Andina, Amazônica, Pacífica, Caribenha, Insular, Orinoco) e mecanismo de pagamento (sem autorização prévia, capitação, pagamento direto, pay-per-case, pay-for-event).

Resultados.

Dos 215.925 casos de pneumonia adquirida na comunidade notificados nos anos 2017-2019, 64,8% ocorreram na região Andina, 73,9% foram pneumonia bacteriana e 1,8% foram pneumonia viral. A adesão às diretrizes foi observada em 5,8% dos casos. Foi maior para crianças com diagnóstico de pneumonia viral (86,0%) em comparação com pneumonia bacteriana (2,0%). Para as crianças com diagnóstico de pneumonia bacteriana, 9,4% receberam algum antibiótico. Uma proporção maior de crianças cobertas por pagamentos capitados (22,3%) recebeu tratamento compatível com as diretrizes, contra apenas 1,3% no esquema de pay-for-event.

Conclusão.

Neste primeiro estudo da Colômbia, a adesão às diretrizes para tratamento ambulatorial de crianças com pneumonia bacteriana foi baixa, sendo melhor para pneumonia viral. Mais estudos qualitativos são necessários para explorar as razões dessa falta de adesão e por qual motivo a pneumonia bacteriana foi a etiologia mais comumente notificada.

Palavras-chave
Pneumonia; criança; pacientes ambulatoriais; amoxicilina; fidelidade a diretrizes; Colômbia

Globally, acute respiratory infection is one of the leading causes of morbidity and mortality, especially among children younger than 5 years (¹1. Leowski J. Mortality from acute respiratory infections in children under 5 years of age: global estimates. World Health Stat Q. 1986;39(2):138–44.). About 15% of episodes of acute respiratory infection are lower respiratory infections (²2. Krishnan A, Amarchand R, Gupta V, Lafond KE, Suliankatchi RA, Saha S, et al. Epidemiology of acute respiratory infections in children-preliminary results of a cohort in a rural north Indian community. BMC Infect Dis. 2015;15(1):1.). Of lower respiratory infections, bronchiolitis and pneumonia are the most frequent, leading to high morbidity and mortality (³3. Guía de práctica clínica para la evaluación del riesgo y manejo inicial de la neumonía en niños y niñas menores de 5 años y bronquiolitis en niños y niñas menores de 2 años [Internet]. Bogotá: Ministerio de de Salud y Protección Social; 2014 p. 207 [cited 2023 Jan 28]. Available from: https://www.minsalud.gov.co/sites/rid/Lists/BibliotecaDigital/RIDE/DE/CA/gpc-completa-neumonia-bronquiolitis.pdf
https://www.minsalud.gov.co/sites/rid/Li... ). In 2019, lower respiratory infections accounted for 3% of mortality in children younger than 5 years in high-income counties, while in South America they accounted for 6%, and in Colombia 10% (⁴4. Global Health Data Exchange [Internet]. Seattle: Institute for Health Metrics and Evaluation, University of Washington; 2021 [cited 2021 May 3]. Available from: https://vizhub.healthdata.org/gbd-compare/
https://vizhub.healthdata.org/gbd-compar... ).

Pneumonia causes more than 900 000 annual infant deaths worldwide (⁵5. Pneumonia. [Internet]. New York, NY: United Nations Children’s Fund; 2022 [cited 2021 May 3]. Available from: https://data.unicef.org/topic/child-health/pneumonia/
https://data.unicef.org/topic/child-heal... ). In 2019, pneumonia was responsible for 14% of all deaths in children younger than 5 years globally (⁶6. Pneumonia in children [Internet]. Geneva: World Health Organization; 2022 [cited 2021 May 3]. Available from: https://www.who.int/news-room/fact-sheets/detail/pneumonia
https://www.who.int/news-room/fact-sheet... ). The most common organisms responsible for pneumonia in children are respiratory viruses (60%), especially respiratory syncytial virus (30%), followed by bacteria (15%), with Streptococcus pneumoniae being the most common (5%) (⁷7. Yun KW, Wallihan R, Juergensen A, Mejias A, Ramilo O. Community-acquired pneumonia in children: myths and facts. Am J Perinatol. 2019;36:S54–7.–⁹9. le Roux DM, Zar HJ. Community-acquired pneumonia in children—a changing spectrum of disease. Pediatr Radiol. 2017;47(11):1392–8.).

In Colombia, the Ministry of Health reports that about 450 children die annually from pneumonia (¹⁰10. Sistema Integrado de Información de la Protección Social [Internet]. Bogotá: Ministerio de Salud y Protección Social; 2021 [cited 2021 Sep 10]. Available from: www.sispro.gov.co
www.sispro.gov.co... ). In 2014, the Ministry published its clinical practice guidelines on risk assessment and management of pneumonia in children younger than 5 years and bronchiolitis in children younger than 2 years (³3. Guía de práctica clínica para la evaluación del riesgo y manejo inicial de la neumonía en niños y niñas menores de 5 años y bronquiolitis en niños y niñas menores de 2 años [Internet]. Bogotá: Ministerio de de Salud y Protección Social; 2014 p. 207 [cited 2023 Jan 28]. Available from: https://www.minsalud.gov.co/sites/rid/Lists/BibliotecaDigital/RIDE/DE/CA/gpc-completa-neumonia-bronquiolitis.pdf
https://www.minsalud.gov.co/sites/rid/Li... ). The aim was to improve the quality of care and optimize the use of resources by managing pneumonia at primary and secondary care levels, while rationalizing antibiotic use (³3. Guía de práctica clínica para la evaluación del riesgo y manejo inicial de la neumonía en niños y niñas menores de 5 años y bronquiolitis en niños y niñas menores de 2 años [Internet]. Bogotá: Ministerio de de Salud y Protección Social; 2014 p. 207 [cited 2023 Jan 28]. Available from: https://www.minsalud.gov.co/sites/rid/Lists/BibliotecaDigital/RIDE/DE/CA/gpc-completa-neumonia-bronquiolitis.pdf
https://www.minsalud.gov.co/sites/rid/Li... ). The Ministry of Health clinical guidelines were in line with the 2014 World Health Organization (WHO) guidance for diagnosis and treatment of childhood pneumonia (¹¹11. Revised WHO Classification and Treatment of Childhood Pneumonia at Health Facilities: Evidence Summaries. Geneva: World Health Organization; 2014:34 [cited 2021 Jul 23]. Available from: https://apps.who.int/iris/handle/10665/137319
https://apps.who.int/iris/handle/10665/1... ).

In children younger than 5 years with community-acquired pneumonia (CAP), not reported as viral pneumonia and managed in an outpatient setting (no hospitalization), the Ministry guidelines recommend empirical treatment with oral amoxicillin for 5 days as first-line therapy. Severe cases of pneumonia must be hospitalized or referred to a higher level of care (³3. Guía de práctica clínica para la evaluación del riesgo y manejo inicial de la neumonía en niños y niñas menores de 5 años y bronquiolitis en niños y niñas menores de 2 años [Internet]. Bogotá: Ministerio de de Salud y Protección Social; 2014 p. 207 [cited 2023 Jan 28]. Available from: https://www.minsalud.gov.co/sites/rid/Lists/BibliotecaDigital/RIDE/DE/CA/gpc-completa-neumonia-bronquiolitis.pdf
https://www.minsalud.gov.co/sites/rid/Li... ). Following management guidelines is key to improving outcomes and guiding appropriate antibiotic use and antibiotic stewardship efforts countrywide. However, adherence to standardized clinical guidance has been recognized as challenging in many settings. Evidence suggests that, in general, compliance with guideline recommendations is limited, including guidelines for chronic diseases, mental health and communicable diseases (¹²12. Docherty M, Shaw K, Goulding L, Parke H, Eassom E, Ali F, et al. Evidence-based guideline implementation in low and middle income countries: lessons for mental health care. Int J Ment Health Syst. 2017;11(1):1–16.–¹⁴14. Cantrell CR, Priest JL, Cook CL, Fincham J, Burch SP. Adherence to treatment guidelines and therapeutic regimens: A US claims-based benchmark of a commercial population. Popul Health Manag. 2011;14(1):33–41.). Non-adherence to guidelines on antibiotic decision-making may result in the irrational or inappropriate prescription of antibiotics, which is an important factor driving antimicrobial resistance (¹⁵15. Global action plan on antimicrobial resistance. Geneva: World Health Organization; 2017:1–28 [cited 2021 Aug 2]. Available from: https://apps.who.int/iris/handle/10665/254352
https://apps.who.int/iris/handle/10665/2... ).

Anecdotal evidence suggests antibiotics other than amoxicillin have been prescribed for non-severe CAP and for patients with a presumptive diagnosis of viral pneumonia in outpatient settings. In Colombia, it has been reported that in children younger than 5 years, 30% of S. pneumoniae isolates outside of the central nervous system are penicillin-resistant, which is partly related to antibiotic overuse (¹⁶16. Vigilancia por laboratorio de aislamientos invasores de Streptococcus pneumoniae Colombia 2006–2018. Vigil por Lab Aisl invasores Streptococcus pneumoniae Colomb 2006-2018 [Internet]. Bogotá: Instituto Nacional de Salud; 2019:1–16 [cited 2021 Nov 18]. Available from: https://www.ins.gov.co/buscador-eventos/Informacindelaboratorio/VigilanciaporlaboratorioS.pneumoniae2006-2018.pdf
https://www.ins.gov.co/buscador-eventos/... ).

Monitoring adherence to Ministry of Health treatment guidelines for CAP is vital to achieving good outcomes, lower mortality, and decreased risk of antimicrobial resistance. The assessment of guideline adherence in clinical treatment can guide stewardship programs and assist with future interventions to support health care providers in Colombia (¹⁷17. Rahman M, Delarosa J, Luies SK, Alom KR, Quintanar-Solares M, Jabeen I, et al. Understanding key drivers and barriers to implementation of the WHO recommendations for the case management of childhood pneumonia and possible serious bacterial infection with amoxicillin dispersible tablets (DT) in Bangladesh: a qualitative study. BMC Health Serv Res. 2020;20(1):1–12.). To date, no assessment of adherence to the Ministry of Health guidance has been made. Therefore. we aimed to: (i) determine the level of adherence to clinical guidelines in prescribing antibiotics in children younger than 5 years with CAP in outpatient settings in Colombia and the factors associated with adherence (age, sex, location, payer source of patients and etiology of pneumonia); and (ii) quantify the type of antibiotics used over time.

METHODS

Study design and data source

This was a cross-sectional study of secondary data extracted from the Integrated Social Protection Information System (SISPRO) of Colombia.

National guidelines for management of CAP in young children

For children younger than 5 years with CAP in outpatient settings, the Ministry of Health guideline recommends that the child should be administered amoxicillin at 90 mg/kg a day (in three oral doses a day) for 5 days.

Severe pneumonia is defined as any child presenting with cough, fever, and rapid breathing, subcostal respiratory retraction, or oxygen saturation less than or equal to 94%. Children with these signs and symptoms should be hospitalized or referred to another institution for hospitalization (³3. Guía de práctica clínica para la evaluación del riesgo y manejo inicial de la neumonía en niños y niñas menores de 5 años y bronquiolitis en niños y niñas menores de 2 años [Internet]. Bogotá: Ministerio de de Salud y Protección Social; 2014 p. 207 [cited 2023 Jan 28]. Available from: https://www.minsalud.gov.co/sites/rid/Lists/BibliotecaDigital/RIDE/DE/CA/gpc-completa-neumonia-bronquiolitis.pdf
https://www.minsalud.gov.co/sites/rid/Li... ).

Health insurance benefits

In Colombia, a collective protection mechanism for the right to health has been established through an insurance scheme and a health benefits plan, financed by the capitation payment unit (¹⁸18. Ministerio de Salud y Protección Social. Resolución 5857 de 2018 [Internet]. Diario Oficial; 2018: 157 [cited 2021 Aug 2]. Available from: https://www.minsalud.gov.co/Normatividad_Nuevo/Resolución5857de2018.pdf
https://www.minsalud.gov.co/Normatividad... ). Under the capitation payment unit, services of contributory and subsidized regimes are covered. The contributory regime covers people who make a monetary contribution to the system because they can pay, while the subsidized regime covers people who cannot pay. Under both regimes, the benefit plan is the same. Consultations, procedures, and medications defined for outpatient management in the clinical practice guide for pneumonia are covered in the health benefits plan (¹⁹19. Regimen contributivo [Internet]. Bogata: Ministerio de Salud y Protección Social; 2016 [cited 2022 May 15]. Available from: https://www.minsalud.gov.co/proteccionsocial/Regimencontributivo/Paginas/regimen-contributivo.aspx
https://www.minsalud.gov.co/proteccionso... ).

SISPRO

The SISPRO database in Colombia is one of the main sources of health information in the country. The database includes anonymized information on patient care, interventions, procedures, medications and costs of all health events reported: health events are coded in accordance with ICD-10 classifications (²⁰20. Solicitud información a entidades promotoras de salud para el “estudio de suficiencia y de los mecanismos de ajuste de riesgo de la unidad de pago por capitación para garantizar el plan de beneficios con cargo a la UPC año 2018.” Bogota: Ministerio de Salud y Protección Social;2017;2:71.). Since the year 2000, all hospitals have been required to make a full report of the services provided (²¹21. Ministerio De Salud resolucion numero 3374 de 2000. Bogota: Ministerio de Salud y Protección Social; 2000:1–6.). Data entry into the SISPRO database is routinely validated.

SISPRO reports the following types of payment mechanisms between the insurer and the hospitals: (i) without prior authorization, when a patient receives a service before the type of agreement or amount is determined and paid; (ii) capitation, advance payment of a fixed sum for a set of activities; iii) direct payment, when a patient receives a service within the insurer-owned hospital network; (iv) pay per case, amount of payment previously agreed for the care of a health condition; and (v) pay for event, an agreed retroactive payment for each service rendered (²⁰20. Solicitud información a entidades promotoras de salud para el “estudio de suficiencia y de los mecanismos de ajuste de riesgo de la unidad de pago por capitación para garantizar el plan de beneficios con cargo a la UPC año 2018.” Bogota: Ministerio de Salud y Protección Social;2017;2:71.).

Study population

We included all children with CAP younger than 5 years reported in SISPRO in Colombia between 2017 and 2019, who had symptoms of non-severe pneumonia and were treated in outpatient facilities.

Variables included

We extracted individual-level data from SISPRO. Variables included were: date of consultation; age (< 1 year, 1–4 years) of child; sex; ICD-10 codes of pneumonia (bacterial, viral, unspecified); region (Andean, Amazonian, Pacific, Caribbean, Insular, Orinoquian); type of health facility (public, private, mixed), level of health facility (I, II, III); payment mechanism (without prior authorization, capitation, direct payment, pay per case, pay for event); amoxicillin prescription (yes, no); and use of amoxicillin with other antibiotics.

For this study, adherence to guidelines was defined as: (i) use of oral amoxicillin if the reported ICD-10 diagnosis was bacterial or pneumonia of unspecified origin; or (ii) use of oral amoxicillin–clavulanate (enzyme inhibitor) after prior use of amoxicillin alone if the reported ICD-10 diagnosis was bacterial or pneumonia of unspecified origin; and (iii) non-use of an antibiotic for pneumonia reported as being of viral etiology.

Statistical analysis

Data were exported in Microsoft Excel 2010 (Microsoft, Redmond, WA, USA) and analyzed using Stata, version 15 (Stata Corporation, College Station, TX, USA). Characteristics of the study sample and antibiotic prescriptions were summarized using proportions and 95% confidence intervals (CI). Antibiotics used (other than amoxicillin) were described by the Anatomical Therapeutic Chemical (ATC) code and the WHO Access Watch Reserve (AWaRe) categories (²²22. Anatomical therapeutic chemical (ATC) classification system [Internet]. Oslo: WHO Collaborating Centre for Drug Statistics Methodology; 2021 [cited 2021 Aug 2]. Available from: https://icdcode.info/atc-ddd/index.html
https://icdcode.info/atc-ddd/index.html... , ²³23. The 2019 WHO AWaRe classification of antibiotics for evaluation and monitoring of use. Geneva: World Health Organization; 2019 [cited 2021 Aug 2]. Available from: https://apps.who.int/iris/handle/10665/327957
https://apps.who.int/iris/handle/10665/3... ).

To determine the association of factors with adherence to guidelines, we used a modified Poisson regression with robust variance estimates. Associations were assessed using adjusted prevalence ratios and 95% CIs. Variables with unadjusted (crude) p-values < 0.2 were included in the model. Collinearity analysis was performed between all potential predictors to determine which variables to include in the multivariable model. For the model development, the Caribbean region was grouped with the Insular region, and the Amazonian region with the Orinoquian region, based on the number of observations from these regions and their proximity to each other.

Ethics

Ethics approval was obtained from the Review Committee of the National Institute of Health of Colombia (no 11/21, date 02/11/2021/), the Ethics Committee of the Pan American Health Organization (no 01/22, date 27/01/2022) and the Ethics Advisory Group of the International Union against Tuberculosis and Lung Diseases, Paris, France (no 09/21, date 02/09/2021). As this study used routinely collected secondary data, we requested a waiver of written informed consent and this was approved by the ethics committees.

RESULTS

Baseline characteristics

A total of 215 925 children younger than 5 years with CAP were included in the study; their baseline characteristics are shown in Table 1. The number of children diagnosed with pneumonia was similar in the 3 years of the study (32.1–34.3%). Children < 1 year of age made up 31.3% of the total and the sex distribution was similar overall. The Andean region reported the most cases (64.8%). The most common reported diagnosis was bacterial pneumonia (73.9%), and most cases were funded by a pay-for-event mechanism (73.2%). More than 90% of the cases had no information on the level of health facility and 67.5% had no data on the type of health facility (public, private, or mixed), so these variables were excluded from the analyses. All cases belonged to the contributory regime.

Use of antibiotics according to etiology of pneumonia

For children diagnosed with bacterial pneumonia, 90.6% were reported as not receiving antibiotic treatment (Table 2). Of those receiving antibiotics, 2.0% received amoxicillin alone. The remaining 7.4% received another antibiotic with or without amoxicillin. When the diagnosis was non-specific pneumonia, amoxicillin treatment alone was reported in 11.3% of cases. For children diagnosed with viral pneumonia, antibiotics were given in 14.4% of cases.

Antibiotics other than amoxicillin

Among the 37 466 cases where antibiotics were reported as being used, 59.9% were of the beta-lactam and penicillin class, while 33.7% were macrolides, lincosamides or streptogramins (Table 3). The database showed that 1557 children were treated with more than one antibiotic and 1337 were prescribed antimicrobials classified in more than one ATC group. No children were reported as being treated with a reserve antibiotic from the WHO AWaRe classification, while 3.3% were treated with more than one antibiotic from the Access and Watch category.

Factors associated with adherence to guidelines

Overall, adherence to guidelines was seen in the treatment of 5.8% of the cases (Table 4). A correlation matrix of all the independent variables did not indicate collinearity between the variables selected for the model. In the multivariable model, probability of adherence decreased in 2018 and 2019. As regards age group, adherence was 57% higher in the older (1–4 years old) versus younger children. Regarding region of residence, the Andean region has the largest population concentration of the entire country (56%) (²⁴24. Informe Observatorio Regional ODS, Región Andina. Bogotá: Observatorio Regional de Objetivos de Desarrollo Sostenible de la Universidad de los Andes; 2020 [cited 2022 May 15]. Available from: https://repositorio.uniandes.edu.co/bitstream/handle/1992/47782/InformeRegiónAndina.pdf?sequence=1&isAllowed=y
https://repositorio.uniandes.edu.co/bits... ), but adherence was 13% lower in the Caribbean/Insular region by than the Andean region. Compared with cases with unspecified pneumonia, adherence to guidelines was 52% lower in cases of bacterial pneumonia, but 8.3 times higher in cases of viral pneumonia. With payment mechanisms, compared with authorized payment, adherence to guidelines was 2.4 times higher in cases using capitation. In contrast, adherence to guidelines was 74% lower in cases using payment per event, compared with authorized payment.

Frequency of use of amoxicillin or another antibiotic

The frequency of use of amoxicillin (+/– clavulanate), clarithromycin or azithromycin by year for treating CAP in children younger than 5 years is shown in Figure 1. Over the three study years, a trend towards non-adherence to guidelines was seen. The use of broad-spectrum antibiotics (clarithromycin, azithromycin, and amoxicillin–clavulanate) increased over time.

DISCUSSION

In this study, we observed a low adherence to clinical guidelines for the management of children with CAP younger than 5 years in outpatient settings in Colombia. While about three quarters of the cases of pneumonia were reported to be of bacterial etiology, adherence to guidelines was better among children diagnosed with viral pneumonia. Capitated payment was associated with higher adherence to treatment guidelines.

Key findings

A high percentage of children were diagnosed with bacterial pneumonia by health care providers. It remains uncertain why providers would report a bacterial diagnosis more frequently than viral, given that around the world pneumonia is mainly caused by viral agents (⁷7. Yun KW, Wallihan R, Juergensen A, Mejias A, Ramilo O. Community-acquired pneumonia in children: myths and facts. Am J Perinatol. 2019;36:S54–7., ⁸8. Ning G, Wang X, Wu D, Yin Z, Li Y, Wang H, et al. The etiology of community-acquired pneumonia among children under 5 years of age in mainland China, 2001–2015: a systematic review. Hum Vaccin Immunother. 2017;13(11):2742–50.). The choice of recording bacterial or unspecified pneumonia rather than viral pneumonia may have something to do with biases in doctors’ classification, mistakes with the providers’ reporting, or difficulties at different levels and types of health facility of which we are unaware (²⁵25. Espinosa O, Rodríguez-Lesmes P, Orozco L, Ávila Di, Enríquez H, Romano G, et al. Estimating cost-effectiveness thresholds under a managed healthcare system: experiences from Colombia. Health Policy Plan. 2022;37(3):359–68.). Challenges in capturing data reported by health providers have been a long-standing problem which is increasing with the greater complexity of medical care (²⁶26. Harper ML, Helmreich RL. Identifying barriers to the success of a reporting system. In: Henriksen K, Battles JB, Marks ES, Lewin DI, editors. Advances in patient safety: from research to implementation (Volume 3: Implementation issues). Rockville, MD: Agency for Healthcare Research and Quality; 2005;167–80., ²⁷27. Ancker JS, Edwards A, Nosal S, Hauser D, Mauer E, Kaushal R. Effects of workload, work complexity, and repeated alerts on alert fatigue in a clinical decision support system. BMC Med Inform Decis Mak. 2017;17(1):1–9.).

Adherence to clinical guidance was low and there were no previous studies on the management of ambulatory CAP in Latin America with which to compare our findings. However, our results are similar to a study in hospitalized CAP cases in Poland (²⁸28. Kraj G, Peradzyńska J, Chądzyńska J, Kulus M, Wołoszyn K, Jackowska T, et al. The influence of national guidelines on the management of community-acquired pneumonia in children. Do pediatricians follow the recommendations? Adv Exp Med Biol. 2019;1211:103–10.). Poor adherence to similar guidelines has been reported in Brazil (26% adherence), while higher adherence has been reported in Pakistan and Lebanon(> 50%) (²⁹29. da Fonseca Lima EJ, Lima DE, Serra GH, Abreu E Lima MA, de Mello MJ. Prescription of antibiotics in community-acquired pneumonia in children: are we following the recommendations? Ther Clin Risk Manag. 2016;12:983–8.–³¹31. Tannous R, Haddad RN, Torbey PH. Management of community-acquired pneumonia in pediatrics: adherence to clinical guidelines. Front Pediatr. 2020;8:1–6.). In fact, antibiotics have shown limited benefits in children younger than 5 years with non-severe pneumonia diagnosed as per WHO criteria (³²32. Lassi ZS, Padhani ZA, Das JK, Salam RA, Bhutta ZA. Antibiotic therapy versus no antibiotic therapy for children aged 2 to 59 months with WHO-defined non-severe pneumonia and wheeze. Cochrane Database Syst Rev. 2021;2021(1).).

We found greater adherence to guidelines in cases with capitated insurance rather than cases covered through payment per event. This finding may be due to the strict auditing and follow-up processes carried out by insurers in the hospitals with which they have the capitation payment mechanism. Similar findings have been reported within the Colombian health system (³³33. Ruiz F, Amaya L, Venegas S. Progressive segmented health insurance: Colombian health reform and access to health services. Health Econ. 2007;16(1):3–18.). Non-capitated services have different operational and financial rules, which could generate a disparity in care for people with similar health conditions.

Strengths

The main strength of this study was that we had direct access to 3 years of SISPRO data, which included diagnostic and treatment information of CAP in children younger than 5 years, for the entire country. This limits the introduction of selection bias, as the data set was countrywide. In addition, we followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (³⁴34. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Int J Surg. 2014;12(12):1495–9.).

Limitations

The main limitation of this study was that the SISPRO databases had incomplete information on variables such as level and type of health facility which are relevant for analyzing diagnosis and antibiotic treatment in relation to different regions of the country. Having this information would assist in future targeted interventions to improve adherence to guidelines.

Another limitation to consider when analyzing our results is that no information was available on comorbidities or vaccination history of the children that could explain the change in antibiotics for the management of non-severe pneumonia. This aspect could be analyzed in other investigations that compile institutional data, medical records and patient cohorts. Finally, as accurate coding of diagnoses is needed for analysis, any shortcomings in the reliability and validity of the SISPRO database could have affected our findings.

Recommendations

First, a rapid qualitative operational research study is needed to examine why providers are diagnosing bacterial CAP more frequently than viral CAP; whether this is because of a wrong clinical diagnosis, unawareness of the diagnostic classification for pneumonia, or incorrect information management by providers. Second, the reason why few data were submitted on level and type of health facility should be assessed to assist decision-makers using the information in SISPRO. Third, health care providers are not closely involved in the review, management and evaluation of information systems and data collection processes. In addition, they do not routinely receive feedback on clinical data quality and monitoring results. Urgent attention is needed to raise awareness of health care providers of the importance of good-quality health data and engage them in the data collection process.

Finally, based on these findings, the need to update the 2014 Colombian pneumonia guideline and include follow-up and adherence mechanisms should be assessed. In addition, an evidence-based educational intervention should be developed for health care providers and managers with a focus on improving adherence to guidelines and data collection on the diagnosis and treatment of CAP. Subsequently, it may be necessary to enhance training activities for clinicians so that they have a better understanding of data reporting and the implications for decision-making. Such measures have the potential to improve health outcomes, save costs and reduce the risk of exacerbation of antimicrobial resistance.

Conclusion

This study found low adherence to guidelines for the management of CAP in children younger than 5 years in outpatient settings in Colombia. Further qualitative systematic studies are required to explore the reasons for this situation, why bacterial pneumonia was the most common etiology reported, and why antibiotics were not prescribed more frequently for bacterial pneumonia. As no previous studies on adherence in Colombia are available, our findings provide baseline data and could help promote improvements in data collection and management for the Colombian health information systems on diagnosis, treatment, monitoring and quality of care.

Disclaimer.

The authors hold sole responsibility for the views expressed in the manuscript, which may not necessarily reflect the opinion or policy of the Revista Panamericana de Salud Pública / Pan American Journal of Public Health and/or those of the Pan American Health Organization.

Acknowledgements.

This research was conducted through the Structured Operational Research and Training Initiative (SORT IT), a global partnership coordinated by the WHO Special Programme for Research and Training in Tropical Diseases (TDR). The specific SORT IT program that led to this publication included a partnership of TDR with the WHO country offices in Colombia and Ecuador and the Pan American Health Organization (PAHO) and was implemented along with The Tuberculosis Research and Prevention Center Non-Governmental Organization, Yerevan, Armenia; The International Union Against Tuberculosis and Lung Diseases, Paris, France and South-East Asia offices, India; Medicins Sans Frontières, Luxembourg, Luxembourg; Medicins Sans Frontières, USA; ICMR–National Institute of Epidemiology, Chennai, India; Sustainable Health Systems, Freetown, Sierra Leone; Institute of Tropical Medicine, Antwerp, Belgium; University of Washington, Seattle, USA, Damien Foundation Brussels, Belgium; Jawaharlal Institute of Postgraduate Medical Education & Research, Pondicherry, India; GMERS Medical College, Gujarat, India; Sri Manakula Vinayagar Medical College, , Madagadipet, India; Universidade Federal de Ciencias da Saude de Porto Alegre, Porto Alegre, Brazil; Ministry of Health, Lima, Peru; Universidade de Brasilia, Brasilia, Brazil; Universidad Central del Ecuador, Quito, Ecuador; Quadram Institute Bioscience, Norwich, , United Kingdom of Great Britain and Northern Ireland; Universidad de los Andes, Bogotá, Columbia; and California State University, Fullerton, USA.

REFERENCES

Publication Dates

History