Assessing the long-term economic impact of wheezing episodes after severe RSV disease in children from Argentina: a cost of illness analysis

Study design

We conducted a prospective cohort study for the evaluation of the COI associated with wheezing episodes that followed a severe RSV-ALRTI within the first year of life in children attended in two public hospitals of Buenos Aires, Argentina. The main study followed for 5 years patients under 12 months of age who were hospitalised due to RSV-ALRTI, allowing for a comprehensive assessment of the long-term economic impact of wheezing syndrome in this population. The analysis was carried out from both public health system and societal perspectives.29 This study adhered to the methodological considerations previously published for studies on the COI.30 31 Patients and the public were not involved in the design, conducting, reporting or dissemination plans of this study.

Data source and study setting

This study employed data derived from a prospective cohort established at two public hospitals located in the southern region of Buenos Aires, Argentina. Both hospitals are comparable in terms of number of inpatient and paediatric intensive care unit (PICU) beds. The population accessing these hospitals is the same. The enrolment process for this cohort took place from May 2014 to August 2016, and the follow-up period was until June 2022. The cohort’s database encompasses details regarding the utilisation of healthcare resources, RSV diagnoses and demographic information of individuals who experienced ALRTI as previously described and detailed in the Instruments section and in online supplemental materials.26

Local context

The healthcare system in Argentina is a mix of public, private and social security providers.32 It is decentralised, with each province managing its own healthcare services.32 The public healthcare system provides services to the entire population, while those with private insurance have access to a network of private clinics and hospitals.32–34 Challenges include disparities in healthcare quality between urban and rural areas and variations in access to specialised services.8 35 In the Buenos Aires Province, approximately 38% of the population does not have health coverage and relies only on the public healthcare system (https://redatam.indec.gob.ar/).36 Patients do not pay any out-of-pocket expenses for their care during their stay: food, diagnostic tests and treatments are all covered by universal health coverage.

Instruments

For the collection of healthcare resources, data collection instruments were developed in REDCap as previously described.26 37 Parents or responsible adults were subject to multiple interviews conducted via telephone calls to gather information regarding their employment status, monthly income and itemised out-of-pocket expenses using research forms and uploading them to REDCap. For long-term wheezing outcomes, the study physicians filled out the research forms on each occasion when patients experienced respiratory exacerbation episodes and attended outpatient health clinics or inpatient paediatric wards. In instances where health visits did not occur, study members maintained periodic contact (every 2 months) with the families through phone calls, WhatsApp messages or email.

Population

Infants aged 12 months or younger hospitalised with severe ALRTI were screened for eligibility. A comprehensive panel of 10 respiratory pathogens, including RSV, was identified in nasopharyngeal aspirates using real-time quantitative PCR (qPCR) in our research laboratory for each eligible patient.6 9 Those infants diagnosed of RSV by qPCR were included in this study.38 Severe ALRTI was defined as the presence of at least one manifestation of lower respiratory tract infection sign (cough, nasal flaring, indrawing of the lower chest wall, subcostal retractions, stridor, rales, rhonchi, wheezing, crackles or crepitations, or observed apnoea) plus hypoxaemia (peripheral oxygen saturation of <95% at room air) or tachypnoea (≥70 breaths per minute from 0 day to 59 days of age, and ≥60 breaths per minute at 60 days of age or older).6 39 40

Infants identified with severe RSV-ALRTI were tracked for a minimum of 5 years to monitor the presence of wheezing episodes (see online supplemental figure 1). The assessment of wheezing episodes was overseen by a study physician or periodically reported by parents through phone calls or face-to-face interviews using the International Study of Asthma and Allergies in Childhood questionnaire.13 41 Inclusion in the cost estimation required the presence of at least one documented episode of wheezing diagnosed by a physician following a severe RSV-ALRTI.

Loss of follow-up was deemed to occur if patients were unreachable through phone, email or WhatsApp, did not visit the primary care centre in the participant hospitals or relocated to another town outside the designated catchment area of the study (Metropolitan Area of Buenos Aires).

Outcome of the study

The study’s objective was to determine the overall cost, considering both direct and indirect costs, associated with hospitalisations and healthcare visits due to wheezing episodes.38 All costs in this study were initially calculated in Argentine pesos (ARS) at the time data collection was obtained and subsequently converted to US dollars using the average exchange rate reported by the Central Bank of Argentina on that specific date.42

Cost estimation

Medical cost estimation was based on the bottom-up costing method described elsewhere.38 In this approach, healthcare resources used were meticulously identified and quantified through the data collection instrument. The unit cost associated with each healthcare resource was extracted from the financial database and the account records of the two participating hospitals.38 It should be noted that since both hospitals are public, the costs are similar as they have standardised case costing procedures.

Through medical visits or phone calls, all 256 patients were followed up for at least 5 years and every wheezing episode was recorded along with the following information:

• Number of visits to the clinician per episode until recovery.

• Severity of the episode according to the Pulmonary Index Score (only for medical visits)43 44: in every visit, presence of wheezing, accessory muscle use, rales and oxygen saturation were documented and then the episode was classified under mild, moderate or severe ALRTI.

• Need for hospitalisation.

For outpatient episodes, if the severity of illness in the visit to the clinician was moderate, a standard treatment on the emergency room (ER) was established that included the administration of albuterol and supplementary oxygen with a Venturi mask, systemic corticosteroids and antipyretic drugs if necessary. The consultation fees for physicians and nurses were also considered in the total cost for outpatients.

For both outpatient and inpatient episodes, the total cost was disaggregated in the following components as previously described38: laboratory tests, labour cost, drugs, feeding, imaging diagnosis, supplies, oxygen supply, overhead and equipment depreciation.38 All these component costs were allocated as previously reported.38 The total cost was calculated per episode and per patient, considering that every subject included may have had one or more episodes of wheezing (with or without hospitalisation requirement).

To estimate the out-of-pocket expenditures, a two-part model was employed to address the substantial proportion of observations at zero combined with skewed positive outcomes (for non-zero).45 A similar approach has been used in a previous study.46 In the first part, three logit models were used to estimate the probability of incurring out-of-pocket expenditures for meals, commute and loss of income, respectively. Briefly, we modelled the natural log odds as a linear function of demographic (sex, age in months, patient’s weight, household income, parent’s education) and clinical (days of hospitalisation, number of wheezing episodes) covariates. Mathematically, we estimated the following equation:

where p is the probability of the outcome of interest y; X and Z are a subset of sociodemographic and clinical covariates; and ε is the error term with distribution   .45

In the second part, a generalised linear model with an identity link function and a gamma distribution for error terms was used to estimate the level of out-of-pocket expenditures for meals, commute and loss of income. This model was chosen instead of the conventional ordinary least squares (OLS) regression because cost data often exhibit significant skewness and adhere to a non-normal distribution, potentially leading to violations of the fundamental assumptions of OLS.47 48 The covariates included in the model were sociodemographic (sex, age in months, patient’s weight, household income, parent’s education) and clinical (days of hospitalisation, number of wheezing episodes). Lastly, the two parts were combined to estimate expected out-of-pocket expenditures by exploiting the basic rule of probability46:

Sample size and statistical analyses

Sample size was calculated using the following formula49:

Precision was established at ±10% of the mean cost, assuming a coefficient of variation of 0.5, a z-score of 1.96 and p (wheezing episodes after severe RSV) of 62%. The minimum sample size necessary for assessing the costs of RSV-associated long-term wheeze was determined to be 115 participants.

We analysed data using Stata (V.16.1, StataCorp, College Station, Texas). Descriptive statistics (frequency and per cent) were used to summarise demographics, clinical variables and healthcare resource utilisation. We applied for the Kolmogorov-Smirnov test to assess the normality of data. For normally distributed variables, we described data using means and SDs; while we described data using the median and IQR for non-normal distributed variables. We described the cost of recurrent wheezing episodes by using the mean together with its 95% CI. We used Student’s t-test to compare the average cost for outpatients and inpatients, and analysis of variance to compare the mean direct medical cost of a wheezing episode between different age groups based on reported wheezing frequency rates (0–12, 12–24 and >24 months of age).11

Study population

Of the initially 470 patients aged 12 months or less screened for eligibility, 256 were confirmed to have severe RSV-ALRTI and enrolled in the study, and then followed for at least 5 years. The mean duration of follow-up for the 256 patients was 48.23 months (95% CI 44.50 to 51.99): 24 patients were lost to follow-up (6 during the 2nd and 3rd years and 18 during the 4th and 5th years). A total of 150 infants (58.6%) experienced at least one episode of wheezing after a severe ALRTI caused by RSV and were included in this analysis. The median number of wheezing episodes per patient reported during the follow-up was 5 (IQR 3–7, range 2–11). The majority of the episodes occurred within 2 years after the first severe RSV-ALRTI: the median time until the last episode of wheezing was recorded is 22 months (IQR 12–33.5).

We compared sociodemographic and clinical-biological variables between patients who developed repeated episodes of wheezing during the first 5 years of life and those who did not (table 1). Although both groups have similar characteristics, patients with no wheezing episodes more frequently lived in precarious housing (built mostly with tin, wood and/or mud) compared with controls (p=0.008). Across the population, the predominant employment status among parents was informal work, with a substantial monthly income variability. The study determined an average household income of US$632.8 per month, leaving more than 70% of the families situated below the poverty line according to Argentina’s income thresholds per family of four members.50

Overall, from the total of 429 episodes of respiratory exacerbations with wheezing, 324 received outpatient treatment, either through visits to the ER or paediatric primary care centres, and 105 needed hospitalisation in paediatric ward. A single wheezing episode may involve multiple health visits throughout its occurrence. No admission to PICUs or fatal events related to episodic wheezing were recorded in this study.

Direct medical costs of wheezing episodes

Considering both outpatient visits and hospitalisations, the mean cost per episode (including all health visits) was calculated at US$185.00 (95% CI 160.33 to 209.67). No PICU admissions were registered for recurrent wheezing episodes. The mean direct medical cost for a severe wheezing episode necessitating hospitalisation, including all health visits before hospital admission, was US$528.32 (95% CI 454.01 to 602.64), as detailed in table 2. The aggregate cost for 105 hospital admissions due to wheezing episodes in 73 patients, requiring 716 hospital bed days (with a median stay of 5 days and an IQR of 3–8 days per episode), was US$60 729.81. In contrast, the average direct medical cost for a single episode not requiring hospital admission was US$85.92 (95% CI 80.51 to 91.34).

When we analysed the cost by patient, the average cost for a hospitalisation was US$544.88 (95% CI 370.67 to 719.09), whereas the cost for an ambulatory health visit was US$58.67 (95% CI 48.13 to 69.22). Finally, we calculated the cumulative direct medical cost per patient, encompassing both severe RSV-associated ALRTI and the total long-term wheezing episodes, resulting in a mean of US$859.44 (95% CI 739.52 to 979.37).38

Similarly, as in the RSV-associated ALRTI COI estimation, labour costs of healthcare workers were a major cost driver of the overall cost (table 2), constituting between 40% and 76% of the total expenses.38

We have analysed each direct medical cost variable, considering age groups across all episodes (table 3), as well as variations in severity (online supplemental tables 1 and 2). We found that the total cost per episode of wheezing was significantly higher (p<0.001) in those infants younger than 12 months of age (207.43, 95% CI 154.3 to 260.6) compared with the subgroups of older toddlers. This difference can be partially explained by higher hospitalisation rates in this age group, leading to increased expenses in laboratory examinations, health worker costs and oxygen supply prices. No significant age-related differences were observed among children hospitalised due to severe obstructive wheezing exacerbations (online supplemental table 1). In contrast, when evaluating the costs associated with outpatient episodes, it became evident that the older subgroup of patients, specifically those aged over 24 months, exhibited a markedly higher total cost (mean $77.35, 95% CI 65.62 to 89.08, p<0.001), along with significantly elevated costs across each comparative variable (online supplemental table 2). Although fewer hospitalisations were recorded in children older than 2 years of age, the number of visits to the paediatrician, the severity of the wheezing episode and, therefore, the need of treatment in the ER before discharge were higher (p<0.001).

Indirect costs and out-of-pocket expenses

Complete data on out-of-pocket expenses and indirect expenses were available for the families of 89 individuals (as shown in table 4). The average non-medical expenses associated with wheezing episodes in the cohort per individual were US$16.36 (95% CI 13.45 to 19.27), and per episode were US$4.89 (95% CI 4.27 to 5.50). The main driver of non-medical costs was food expenditures. Expenses were significantly higher in those hospitalised patients.

Total cost related to long-term wheezing episodes

Per wheezing episode, the total cost including medical and non-medical expenses was US$191.01 (95% CI 166.37 to 215.64). Based on the level of care, we observed that the total cost for hospitalisations was significantly higher (US$535.55, 95% CI 461.25 to 609.84) compared with outpatient episodes (US$92.34, 95% CI 86.74 to 97.94). In table 5, we presented the cumulative cost per individual, encompassing all wheezing episodes with and without the inclusion of acute RSV hospitalisations. Interestingly, the average total cost of all the wheezing episodes per child was US$415.99, with a maximum cost of US$5391.96, while adding RSV hospitalisation increased the cost to US$959.56 with a maximum of US$6024.35.