Secondary Logo

Journal Logo

Hospitalization trends, costs, and risk factors in HIV-infected children on antiretroviral therapy

Collins, Intira J.a,b,c,d; Cairns, Johne; Jourdain, Gonzaguea,b,c; Fregonese, Federicaf; Nantarukchaikul, Maneeratng; Lertpienthum, Narongh; Wannarit, Pornpuni; Attavinijtrakarn, Pornsawanj; Layangool, Prapaisrik; Le Coeur, Sophiea,b,c,l; Lallemant, Marca,b,cfor the Program for HIV Prevention and Treatment (PHPT) study team

doi: 10.1097/QAD.0b013e328357f7b9
EPIDEMIOLOGY AND SOCIAL
Free
SDC

Objective: To assess hospitalization trends in HIV-infected children on antiretroviral therapy (ART) in Thailand, an important indicator of morbidity, ART effectiveness, and health service utilization.

Design: Prospective observational cohort

Method: Children initiating ART in 1999–2009 were followed in 40 public hospitals. Hospitalization rate per 100 person-years were calculated from ART initiation to last follow-up/death. Costs to the healthcare provider were calculated using WHO inpatient estimates for Thailand. Zero-inflated Poisson models were used to examine risk factors for early (<12 months of ART) and late hospitalization (≥12 months) and frequency of admissions.

Results: A total of 578 children initiated ART, median follow-up being 64 months [interquartile range (IQR) 43–82]; 211 (37%) children were hospitalized with 451 admissions. Hospitalization rates declined from 63 per 100 person-years at less than 6 months to approximately 10 per 100 person-years after 2 years of ART, and costs fell from $35 per patient-month to under $5, respectively. Age less than 2 years, US Centers of Disease Control and Prevention stage B/C, and stunting at ART initiation were associated with early hospitalization. Among those hospitalized, baseline CD4 cell percentage less than 5%, wasting, initiation on dual therapy, late calendar year, and female sex were associated with higher incidence of early admissions (P <0.02). There were no predictors of late hospitalization, although previous hospitalization in less than 12 months of ART was associated with three times higher incidence of late admissions (P < 0.0001).

Conclusion: One in three children required hospitalization after ART. Admissions were highest in the first year of therapy and rapidly declined thereafter. Young age, advanced disease stage, and stunting at baseline were predictive of early hospitalization. Treatment initiation before disease progression would likely reduce hospitalization and alleviate demands on healthcare services.

Supplemental Digital Content is available in the text

aInstitut de Recherche pour le Dévecnt (IRD URI 174), Marseille, France

bFaculty of Associated Medical Sciences, Chiang Mai University, Chiang Mai, Thailand

cDepartment of Immunology and Infectious Diseases, Harvard School of Public Health, Boston, Massachusetts, USA

dFaculty of Epidemiology and Population Health

eFaculty of Public Health and Policy, London School of Hygiene & Tropical Medicine, London, UK

fPadova University, Padua, Italy

gSomdej Prapinklao Hospital, Bangkok

hBuddhachinaraj Hospital, Pitsanuloke

iLamphun Provincial Hospital, Lamphun

jPhaholpolpayuhasaena Hospital, Kajanaburi

kBhumibol Adulyadej Hospital, Bangkok, Thailand

lUnité Mixte de Recherche 196 Centre Français de la Population et du Développement (INED-IRD-Paris V Université), Paris, France.

Correspondence to Intira Collins, London School of Hygiene & Tropical Medicine, Keppel Street, London WC1E 7HT, UK. Tel: +44 207 636 8636; fax: +44 207 436 5389; e-mail: intira.collins@lshtm.ac.uk

Received 13 April, 2012

Revised 27 June, 2012

Accepted 10 July, 2012

The study was presented in part in 3rd International Pediatric AIDS Conference, Rome, Italy, 15–16 July 2011 (abstract P_43).

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Website (http://www.AIDSonline.com).

Back to Top | Article Outline

Introduction

Several studies have reported dramatic reductions in mortality and hospitalization among HIV-infected children following the introduction of antiretroviral therapy (ART) [1–4]. However, the long-term trends in hospitalization of children on ART are less well described, particularly in resource-limited settings in which children often initiate therapy at advanced disease stage [5–7]. As HIV-infected children survive into adolescence, hospitalization represents an important outcome measure, as an indicator of morbidity and healthcare utilization. Furthermore, there are scant data on hospitalization costs from the healthcare provider's perspective – essential to inform program financing and decision analytic models for optimal allocation of resources [8].

In this study, we analyzed data from a large prospective observational treatment cohort of HIV-infected children in Thailand (Clinicaltrials.gov NCT00433030). The objectives were to describe the rates of hospitalization, their primary causes, outcomes, and cost over time on ART. We assessed risk factors of hospitalization and admission frequency. We hypothesized that children who initiated ART at advanced disease stage were at higher risk of hospitalization and repeated admissions.

Back to Top | Article Outline

Methods

Study population

HIV-infected children received ART in a network of 40 public hospitals as part of the national scale-up efforts, as described elsewhere [9]. Children aged less than 18 years, antiretroviral naive (except prophylaxes for prevention of mother-to-child transmission of HIV) who initiated ART (defined as ≥2 drugs) between January 1999 and January 2009, were included in this analysis. ART initiation was based upon clinical and immune criteria: US Centers of Disease Control and Prevention (CDC) HIV clinical disease stage B or C or CD4 T-cell percentage less than 20% if less than 2 years of age, and CD4 less than 15% if 2 years of age or above [10,11]. ART and laboratory monitoring were provided free of charge. Parents/guardians provided written informed consent at study entry, and from December 2006, assent was requested from children aged 8 years or above. The study was approved by the Thai Ministry of Public Health and local Ethics Committees.

Back to Top | Article Outline

Hospitalization data

At monthly visits, nurses/physicians asked the child/caregiver about the child's well being, including any hospitalizations in the previous month at the site or any other hospital. All serious adverse events (SAEs), defined as life threatening, requiring/prolonging hospitalization, or resulting in disability or death [12], were reported by the attending physician. If hospitalization occurred in another hospital, the nurse/physician interviewed the caregiver for the details.

All SAEs involving hospitalization, defined as admission for inpatient care for any duration and cause, were included in this analysis. The following data were extracted (through 31 August 2010): date of admission and discharge, primary cause of hospitalization, and outcome. The primary causes were categorized based on an organ and disease system coding scheme; diseases not within these categories were coded as ‘other’ (Supplementary Table 1, http://links.lww.com/QAD/A242). One child may contribute multiple events. Hospitalizations prior to ART initiation were excluded as they represented patients in a natural history state. When children initiated ART during hospitalization, the admission date was set at start of therapy (11 cases). Children who were discharged or died on the date of admission were assigned a stay of 0.5 days.

Costs of hospitalization were based on WHO unit cost estimate of inpatient care per day in Thailand, assuming an 80% occupancy rate, excluding drug costs, and expressed in international dollars (Int$ 2000): $47.13, $61.38, and $83.71 per day in primary, secondary, and tertiary hospitals, respectively [13]. These cost estimates were based on an extensive literature review and modeling inpatient cost as a ratio of outpatient visit costs. The use of international dollars, based on purchasing power parity exchange rates, is recommended for nontraded goods such as inpatient day to allow for comparisons across settings [14]. The 40 hospitals in this study were categorized by specialist services provided and bed capacity [(14 primary, 15 secondary, and 11 tertiary hospitals). Primary hospital is defined as few clinical specialities with 30–200 beds; secondary hospital: 5–10 clinical specialities with 200–800 beds; tertiary hospital: highly specialized staff and equipment and may have teaching activities with 300–1500 beds] and corresponding costs were applied [13].

Back to Top | Article Outline

Statistical analyses

Number of hospital admissions per 100 person-years of observation was calculated for the overall period and by 6-month and 12-month intervals. Children were at risk from start of ART to last follow-up or death.

Cost per hospitalization was the duration of admission multiplied by daily cost of inpatient care according to the hospital category. The mean and total cost of hospitalization was calculated per 6-month interval, and the cost per patient-month calculated based on total cost divided by person-month of follow-up.

Back to Top | Article Outline

Risk factors of hospitalization

The outcome of interest was number of hospitalizations per child during two time periods: less than 12 months after start of ART (early hospitalization) and over 12 months (late hospitalization). This cut off was based on the median time to all admissions and previous studies showing markedly different hospitalization rates before and after 12 months of therapy [15].

Potential risk factors for early and late hospitalization were baseline characteristics at start of ART with less than 20% of missing data: sex, age, CD4%, viral load, CDC disease stage, weight-for-age z-score (WAZ), height-for-age z-score (HAZ) and weight-for-height for age z-score (WHZ) based on Thai reference curves [16,17], initial regimen, calendar year of initiation, and duration of follow-up. All baseline characteristics included in the analysis had less than 10% missing data, except for viral load (16% missing). Baseline hemoglobin was excluded due to excess missing (26%); WAZ was correlated to WHZ and, therefore, only the latter was included. Analysis of late hospitalization was restricted to children on follow-up at 12 months and includes early hospitalization as a potential risk factor.

Back to Top | Article Outline

Regression models

More than 50% of children were never hospitalized resulting in a large proportion with an outcome of zero, thus violating the normal distribution assumptions of standard regression models [18]. Zero-inflated models address this issue and are increasingly used for analysis of health resource utilization [19–23]. We compared four regression models: Poisson, negative binomial, zero-inflated Poisson (ZIP), and zero-inflated negative binomial (ZINB) model. Residuals of the predicted versus observed values were compared using the Long and Freese method to assess the best fit [24].

The ZIP model was chosen for the final model, with the best fit to the observed data [The ZIP model was preferred due to over dispersion of data and preponderance of zero events: 77% of children were never hospitalized in the first 12 months of ART, and 77% never hospitalized after 12 months (Vuong test P <0.001 for ZIP versus Poisson model). The ZINB model was unstable with high standard errors and not pursued (data not shown).] (Supplementary Figure 1, http://links.lww.com/QAD/A242). The ZIP model addresses two questions simultaneously; first, what predicts the dichotomous outcome of never versus ever hospitalized? To ease interpretation, coefficients were transformed into odd ratios of being hospitalized. Second, among hospitalized children, what predicts the frequency of admissions? The latter is expressed as incidence rate ratios (IRRs). In univariable analysis, variables associated with each outcome at P value less than 0.2 were included in their respective components of the multivariable analyses using backward selection, and associated variables with P value 0.2 or less were kept in the final model. Data were analyzed using STATA version 11 (Stata Corp., College Station, Texas, USA).

Back to Top | Article Outline

Results

A total of 578 children were included in the analysis. At ART initiation, the median (interquartile range, IQR) age was 7 years (2–10), CD4% 7% (2–16) and 52% of children were in CDC disease stage B or C (Table 1). The median duration of follow-up was 64 months (IQR, 43–82), with a total of 2918 person-years of follow-up.

Table 1

Table 1

Overall 211 (37%) children were hospitalized with 451 separate admissions (Table 2). Their median time to first admission was 4 months (IQR, 1–22) after start of ART. Among hospitalized children, 116 (55%) had one admission, 49 (23%) had two, 22 (10%) had three, and 24 (11%) had four or more (range 1–23). The mean duration of hospitalization was 8.6 days (SD 10.5 days). Outcomes were documented in all but one event: 400 (89%) were resolved, 15 (3%) voluntary discharged, 23 (5%) died during hospitalization, and 12 (3%) were resolved/discharged but died within 3 months.

Table 2

Table 2

Back to Top | Article Outline

Hospitalization rates, costs, and primary cause

The overall hospitalization rate was 15.4 per 100 person-years. Rates varied significantly with time: 62.9 per 100 person-years at less than 6 months of ART; 22.8 at 6–12 months; and stabilized around 10 hospitalization per 100 person-years after 2 years (Table 2, Fig. 1a). Thirty-eight percent (n = 172) of hospitalizations occurred within the first 6 months of therapy, and 51% (n = 231) within the first year. In the first 6 months, one of the highest hospitalization rates was in children with baseline CD4 cell percentage less than 5%: 90 per 100 person-years [95% confidence interval (CI) 73.3–110.3] as compared to 49 (95% CI 35.7–68.1) per 100 person-years in children with baseline CD4 cell percentage of 15% and above. However, after 6 months, the rates began to converge and by 18 months, children with baseline CD4 cell percentage less than 5% had less than 20 admissions per 100 person-years, comparable to children with higher baseline CD4% (Fig. 1b).

Fig. 1

Fig. 1

The mean cost of hospitalization was $580 per admission (SD 726, range $24–5902), although this varied with time, peaking at $650–750 during the first 18 months on ART and gradually declining thereafter (Table 2). The hospitalization cost for the cohort declined from $35 per patient-month during the first 6 months of ART to less than $5 per patient-month after 2 years (Fig. 1a).

The primary cause of hospitalization was infectious disease-related in 317 (71%) admissions, with pneumonia as the leading cause accounting for 30% of all admissions (Table 3). Noninfectious causes accounted for 94 (21%) admissions, including 15 (2.6%) for cardiovascular diseases. Ten admissions (2%) were classified as antiretroviral drug related (Supplementary Table 1, http://links.lww.com/QAD/A242). The primary causes of hospitalization were similar before and after 12 months of ART.

Table 3

Table 3

Back to Top | Article Outline

Risk factors of early and late hospitalization

Overall 133 children (23%) experienced early hospitalization in the first 12 months of ART with 231 admissions (range 1–8). Early hospitalization was independently associated with young age, stunting, and advanced disease stage at ART initiation. Children aged less than 2 years had five times the odds of early hospitalization as compared to those aged 8 years and over (P = 0.007); CDC stage B/C had two times the odds of early hospitalization as compared to stage N/A (P = 0.05); and low baseline HAZ had 1.4 odds of early hospitalization per z-score decrease (P = 0.001; Table 4). Among children with early hospitalization, CD4 cell percentage less than 5% and low WHZ at ART initiation were independently associated with higher incidence of early admissions, as well as initiation on dual therapy, late calendar year, and female sex (all P < 0.02).

Table 4

Table 4

At 12 months of ART, 511 (88%) children were on follow-up (31 died, 17 lost to follow-up, and 19 transferred out of the study) and included in analysis of late hospitalization: 115 (22.5%) were hospitalized with 220 admissions (range 1–21). There was no significant predictor of late hospitalization. However, among children hospitalized, higher incidence of late admissions was associated with low baseline WHZ (IRR 1.2 admissions per z-score decrease, P = 0.004), and early hospitalization (IRR 3.5, P < 0.001). Conversely, baseline age 2–7 years was associated with lower incidence of late admissions as compared to children aged 8 years and over (P < 0.0001). There appeared to be an effect of efavirenz-based regimen, but this was no longer observed when two outlier cases (with 11 and 21 late hospitalizations) were excluded (P = 0.15).

Back to Top | Article Outline

Discussion

One in three children required hospitalization in 5 years of follow-up on ART. The overall hospitalization rate was 15.4 per 100 person-years, but varied with time, peaking at 63 per 100 person-years at 6 months and falling to approximately 10 per 100 person-years after 2 years of ART. Similar findings were reported in a smaller cohort of 129 children in northern Thailand, with 35% of children hospitalized, mostly during the first 6 months of therapy. Their overall hospitalization rate of 23.0 per 100 person-years was higher than that in our cohort, most likely due to their shorter follow-up time of 2 years [15]. Recent studies in Africa and India have been restricted to cause-specific hospitalizations or focused on HIV-exposed rather than infected children and, therefore, were not comparable [25,26].

Studies in the United States and western Europe have reported extremely low hospitalization rates of less than 1 admission per 100 person-years in the HAART era [1,2,27,28]. However, those estimates were based on hospitalizations per calendar year rather than time since start of therapy and included a mixed population of children on HAART and those not yet requiring treatment. Furthermore, the treated children often initiated therapy during infancy prior to disease progression.

Studies in adults on ART have reported a decline in infectious disease-related hospitalizations accompanied by a rise in non-AIDS-related admissions over time [29,30]. In contrast, pediatric studies in Italy and Thailand suggest that infections, particularly pneumonia and bacterial infections, still constitute a large proportion of hospitalizations [2,15]. Indeed, in our study, 71% of hospitalizations were due to infections, mostly pneumonia and this proportion did not decrease even after 12 months of ART. Ten admissions were reported as antiretroviral drug-related, although 43 admissions due to hematological diseases (mainly anemia) and six admissions for metabolic diseases could have been partly antiretroviral drug-related. With only one admission due to immune reconstitution syndrome (IRIS), the IRIS incidence was lower than that in previous studies [15], most likely due to our classification based on the underlying illness.

Previous studies suggest that hospitalization constitute 15–49% of total cost of HIV care with highest costs incurred in the first year of treatment and/or patients at most advanced disease stage [4,31,32]. Our findings are consistent with those reports with highest costs peaking at $35 per patient-month during the first 6 months of therapy and high incidence of early hospitalization was associated with advanced disease progression at initiation. Hospitalization cost then declined to less than $5 after 2 years of ART, and less than $2 per patient-month after 4 years. This suggests that children on ART incur very low inpatient care costs as treatment programs mature, reducing the burden on healthcare services and releasing resources for other competing illnesses.

There are few costing studies to compare with. Studies in resource-limited setting have been restricted to adults and do not report disaggregated hospitalization costs [32,33]. Two pediatric studies, both in the United States, reported hospitalization costs of approximately $180 per patient-month in the HAART era [4,31]. This is over five times our highest cost estimate, despite the US cohorts reporting very low hospitalization rates of 0.28 admissions per 100 person-years [31] and shorter duration of admission [4], suggesting their costs were mainly driven by higher cost of inpatient care.

We assessed risk factors for early and late hospitalization using the ZIP model, which allows for the excess number of children with zero events. The model also utilizes the richness of the available data, with analysis of predictors of hospitalization as well as frequency of admissions. This approach departs from previous studies using only the binary outcome of ever versus never hospitalized, which fails to distinguish patients with single versus multiple admissions and the potential differences in risk factors [2,27,34].

In our cohort, children initiating ART at less than 2 years were at highest risk of early hospitalization. Similar trends have been reported in Europe and the United States [1,27], and reflects the high morbidity and mortality during infancy, particularly when therapy was initiated when immunocompromised [9,35]. CDC disease stage B/C and stunting at baseline were independent predictors of early hospitalization, whereas CD4% was not. However, among hospitalized children, low baseline CD4% and wasting were associated with higher incidence of early admissions. These findings support WHO recommendations for immediate initiation of therapy in young children less than 2 years and timely initiation of therapy in older children prior to advanced disease progression [36].

Initiation on dual therapy, late calendar year, and female sex were also associated with frequent early admissions. The effect of dual therapy may reflect suboptimal viral suppression associated with such combinations before availability of HAART. The effect of late calendar year of initiation may reflect increased experience of healthcare providers more readily admitting children with complications. It is unclear why female sex was associated with higher incidence of early hospitalization, although this trend has been reported in adult studies [29,32].

There were no predictors of late hospitalization, suggesting that children who survive the first year of therapy have similar risk of long-term hospitalization regardless of their baseline characteristics. However, among those hospitalized, children with early hospitalization had 3.5 times higher incidence of late admissions, which may reflect vulnerability to health problems or unresolved underlying conditions.

These results were comparable with the Poisson and negative binomial models, although those models did not differentiate predictors of hospitalization as compared to frequency of admissions; therefore, the effect of some associated variables was diluted as compared to the ZIP model (data not shown).

There are several study limitations to consider. First, hospitalizations occurring outside the study sites may have been underreported, although this is likely to be rare due to the monthly visits and free provision of care. However, our cohort has reported lower mortality rates as compared to the national treatment program [6,9], most likely due to the close monitoring and experienced care team. It is unclear whether these factors would also reduce the risk of hospitalization. Importantly, the proportion of children hospitalized was very similar to the northern Thailand cohort [15], the characteristics of children at start of ART are comparable to the national program [6], and the HIV care teams also implement the national program at their sites, and, therefore, the benefits of their experience would extend beyond our cohort.

Second, hospitalization costs were calculated based on WHO generic unit cost of inpatient care in Thailand because patient-level cost data were not available. The WHO unit costs are not specific to HIV disease or a pediatric population but do represent a standardized estimate, which allows for comparisons across populations and diseases. Third, the diagnosed causes of hospitalization may include misclassification, in particular those related to IRIS or ART toxicities. Fourth, risk factors of hospitalization were limited to baseline characteristics at start of ART, whereas time-updated variables such as current treatment, CD4%, and viral load were not included; the latter two have been reported as predictive of hospitalization in cross-sectional studies [2,27].

Back to Top | Article Outline

Conclusion

In summary, one in three children was hospitalized on ART, with highest risk in the first year of therapy. Young age and clinical indicators of advanced disease progression at start of ART were associated with early hospitalization, frequent admissions, and consequently higher costs to the healthcare system. Implementation of the WHO revised guidelines for immediate therapy in HIV-infected infants less than 2 years regardless of clinical or immune status [37,38], is likely to reduce mortality and hospitalization among the youngest children. Further studies on hospitalization trends and costs are warranted to evaluate the impact of the new guidelines, forecast the long-term costs of ART provision, and assess the evolving needs of HIV-infected children entering into adolescents and adulthood.

Back to Top | Article Outline

Acknowledgements

The authors would like to thank all the children, their families, and all members of the hospital teams of the participating sites involved in the PHPT pediatric cohort study.

Participating sites and principal investigators are as follows:

Chiangrai Prachanukroh: R. Hansudewechakul, K. Preedisripipat, C. Chanta; Nakornping: S. Kanjanavanit; Prapokklao: C. Ngampiyaskul, N. Srisawasdi; Chonburi: S. Hongsiriwan; Bhumibol Adulyadej: P. Layangool, J. Mekmullica; Phayao Provincial: P. Techakunakorn, S. Sriminiphant; Samutsakhon: P. Thanasiri, S. Krikajornkitti; Kalasin: S. Srirojana; Lamphun: P. Wannarit, K. Pagdi, R. Kosonsasitorn, R. Somsamai; Mae Chan: S. Buranabanjasatean; Sanpatong: N. Akarathum; Somdej Prapinklao: N. Kamonpakorn, M. Nantarukchaikul; Phan: S. Jungpichanvanich; Phaholpolphayuhasena: P. Attavinijtrakarn; Samutprakarn: A. Puangsombat, C. Sriwacharakarn; Rayong: W. Karnchanamayul; Buddhachinaraj: N. Lertpienthum, W. Ardong; Nakhonpathom: S. Bunjongpak; Health Promotion Region 6 Khon Kaen: S. Hanpinitsak, N. Pramukkul; Somdej Pranangchao Sirikit: T. Hinjiranandana; Mae Sai: S. Kunkongkapan; Chacheongsao: R. Kwanchaipanich; Chiang Kham: V. Wanchaitanawong, P. Jittamala; Pranangklao: P. Lucksanapisitkul, S. Watanayothin; Health Promotion Region 10 Chiang Mai: W. Jitphiankha, K. Jittayanun; Hat Yai: B. Warachit, T. Borkird; Mahasarakam: S. Na-Rajsima, K. Kovitanggoon; Ratchaburi: C. Sutthipong, O. Bamroongshawkaseme.

PHPT Clinical Trial unit

Sites monitoring: P. Sukrakanchana, S. Chalermpantmetagul, C. Kanabkaew, R. Peongjakta, J. Chaiwan, Y. S. Thammajitsagul, R. Wongchai, N. Kruenual, N. Krapunpongsakul, W. Pongchaisit, T. Thimakam, R. Kaewsai, J. Wallapachai, J Thonglo, S. Jinasa, J Khanmali, P. Chart, J. Chalasin, B. Ratchanee, N. Thuenyeanyong, P. Krueduangkam, P. Thuraset, S. Thongsuwan, W. Khamjakkaew; Laboratory: P. Tungyai, J. Kamkorn, W. Pilonpongsathorn, P. Pongpunyayuen, P. Mongkolwat, L. Laomanit, N. Wangsaeng, S. Surajinda, W. Danpaiboon, Y. Taworn, D. Saeng-ai, A. Kaewbundit, A. Khanpanya, N. Boonpleum, P. Sothanapaisan, P. Punyathi, P. Khantarag, R. Dusadeepong, T. Donchai, U. Tungchittrapituk, W. Sripaoraya; PHPT Data center: S. Tanasri, S. Chailert, R. Seubmongkolchai, A. Wongja, K. Yoddee, K. Chaokasem, P. Chailert, K. Suebmongkolchai, A. Seubmongkolchai, C. Chimplee, K. Saopang, P. Chusut, S. Suekrasae, T. Yaowarat, B. Thongpunchang, T. Chitkawin, A. Lueanyod, D. Jianphinitnan, J. Inkom, N. Naratee, N. Homkham, T. Thasit, W. Wongwai, W. Chanthaweethip, R. Suaysod, T. Vorapongpisan, N. Jaisieng; Administrative support: N. Chaiboonruang, P. Pirom, T. Thaiyanant, T. Intaboonma; S. Jitharidkul, S. Jaisook, D. Punyatiam, L. Summanuch, N. Rawanchaikul, P. Palidta, S. Nupradit, T. Tankool, W. Champa; Tracking & Supplies: K. Than-in-at, M. Inta, R. Wongsang; D. Chinwong, C.Sanjoom, P. Saenchitta, P. Wimolwattanasarn, N. Mungkhala,

I.J.C. contributed to the study design, conducted the analyses, writing and revision of the article. I.J.C., J.C., G.J., F.F., and M.L. contributed to the study design, interpretation of data, and critical revisions of the article. M.N., N.L., P.W., P.A., P.L., and S.L. contributed to the study design, data collection, and critical revisions of the article.

The present work was supported by the Global Fund to fight AIDS, Tuberculosis and Malaria (Thailand Grant Round 1 sub recipient PR-A-N-008); Ministry of Public Health, Thailand; Oxfam Great Britain, Thailand; Institut de Recherche pour le Développement (IRD), France; Institut National d’Etudes Démographiques, France; Department of Technical and Economic Cooperation, Thai International Cooperation Agency; International Maternal Pediatric Adolescents Aids Clinical Trials Group (IMPAACT), The National Institutes of Health, US (R01 HD 33326; R01 HD 39615). United Kingdom Medical Research Council Doctoral Training Account Studentship for I.C.

Back to Top | Article Outline

Conflicts of interest

There are no conflicts of interest.

Back to Top | Article Outline

References

1. Gibb DM, Duong T, Tookey PA, Sharland M, Tudor-Williams G, Novelli V, et al.Decline in mortality, AIDS, and hospital admissions in perinatally HIV-1 infected children in the United Kingdom and Ireland. BMJ 2003; 327:1019–1025.
2. Viani RM, Araneta MRG, Deville JG, Spector SA. Decrease in hospitalization and mortality rates among children with perinatally acquired HIV type 1 infection receiving highly active antiretroviral therapy. Clin Infect Dis 2004; 39:725–731.
3. Bertolli J, Hsu HW, Sukalac T, Williamson J, Peters V, Frederick T, et al.; Pediatric Spectrum of HIV Disease Study. Hospitalization trends among children and youths with perinatal human immunodeficiency virus infection, 1990–2002. Pediatr Infect Dis J 2006; 25:628–633.
4. Wilson LS, Basu R, Christenson M, Hensic L, Paoli C, Wara D, et al. Pediatric HIV costs across three treatment eras from 1986 to 2007. Pediatrics 2010; 126:e541–e549.
5. Sutcliffe CG, van Dijk JH, Bolton C, Persaud D, Moss WJ. Effectiveness of antiretroviral therapy among HIV-infected children in sub-Saharan Africa. Lancet Infect Dis 2008; 8:477–489.
6. McConnell MC, Chasombat S, Siangphoe U, Yuktanont P, Lolekha R, Pattarapayoon N, et al. National Program Scale-Up and Patient Outcomes in a Pediatric Antiretroviral Treatment Program, Thailand, 2000-2007. J Acquir Immune Defic Syndr 2010; 54:423–429.
7. Peacock-Villada E, Richardson B, John-Stewart G. Paediatric HAART treatment outcomes: a comparison of developing and developed countries. 16th Conference on Retroviruses and Opportunistic Infections (CROI 2009); Montreal, Canada; 8–11 February 2009.
8. Levy A, James D, Johnston K, Hogg R, Harrigan P, Harrigan B, et al. Direct costs of HIV/AIDS care. Lancet Infect Dis 2006; 6:171–177.
9. Collins I, Jourdain G, Hansudewechakul R, Kanjanavanit S, Hongsiriwon S, Ngampiyasakul C, et al. Long term survival of HIV infected children receiving antiretroviral therapy in thailand: a 5 year observational cohort study. Clin Infect Dis 2010; 51:1449–1457.
10. Working Group on Antiretroviral Therapy and Medical Management in HIV-Infected Children. Guidelines for the use of antiretroviral agents in pediatric HIV infection, 7 January 2000. http://http://www.aidsinfo.nih.gov/contentfiles/PediatricGuidelines01072000023.pdf. [Accessed 6 August 2012]
11. Thai Ministry of Public Health. National guidelines for the clinical management of HIV infection in children and adults. 7th ed. Bangkok; 2002.
12. International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use Guidelines Clinical Safety Data Management: definitions and standards for expedited reporting E2A; 1994. http://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Efficacy/E2A/Step4/E2A_Guideline.pdf [Accessed 16 May 2011].
13. Adam T, Evans DB. Determinants of variation in the cost of inpatient stays versus outpatient visits in hospitals: a multicountry analysis. Soc Sci Med 2006; 63:1700–1710.
14. World Health Organization. Making choices in health: WHO guide to cost-effectiveness analysis. In: Tan-Torres Edejer T, Baltussen R, Adam T, Hutubessy R, Acharya A, Evans DB, Murray CJL, editors; 2003. http://www.who.int/choice/publications/p_2003_generalised_cea.pdf. [Accessed 16 May 2011].
15. Puthanakit T, Aurpibul L, Oberdorfer P, Akarathum N, Kanjananit S, Wannarit P, et al. Hospitalization and mortality among HIV-infected children after receiving highly active antiretroviral therapy. Clin Infect Dis 2007; 44:599–604.
16. Working Group on using Weight and Height References in Evaluating the Growth Status of Thai Children. Manual on using weight and height references in evaluating the growth status of Thai children. Bangkok: Department of Health, Ministry of Public Health; 2000.
17. Traisathit P, Jourdain G, Le Coeur S, Briand N, Lechenadec J, Mekmullica J, et al. Effect of antiretroviral therapy on the growth of HIV infected children in Thailand. 1st International Workshop on Pediatric HIV; Cape Town, South Africa; 2009 (Poster 40).
18. Jones A. Applied econometrics for health economists: a practical guide. 2nd ed. Radcliffe Publishing Ltd.; 2007.
19. Logan DE, Radcliffe J, Smith-Whitley K. Parent factors and adolescent sickle cell disease: associations with patterns of health service use. J Pediatr Psychol 2002; 27:475–484.
20. Karazsia BT, van Dulmen MH. Regression models for count data: illustrations using longitudinal predictors of childhood injury. J Pediatr Psychol 2008; 33:1076–1084.
21. Zaninotto P, Falaschetti E. Comparison of methods for modelling a count outcome with excess zeros: application to activities of daily living (ADL-s). J Epidemiol Community Health 2011; 65:205–210.
22. Cheung YB. Zero-inflated models for regression analysis of count data: a study of growth and development. Stat Med 2002; 21:1461–1469.
23. Moon S, Shin J. Healthcare utilization among Medicare-Medicaid dual eligibles: a count data analysis. BMC Public Health 2006; 6:88.
24. Long JS, Freese J. Countfit: commands for the post-estimation interpretation of regression models. http://www.ats.ucla.edu/stat/stata/faq/countfit.htm.[Accessed10February2011].
25. Prendergast A, Bwakura-Dangarembizi M, Cook AD, Bakeera-Kitaka S, Natukunda E, Nahirya Ntege P, et al. Hospitalization for severe malnutrition among HIV-infected children starting antiretroviral therapy. AIDS 2011; 25:951–956.
26. Singh H, Gupte N, Kinikar A, Bharadwaj R, Sastry J, Suryavanshi N, et al. High rates of all-cause and gastroenteritis-related hospitalization morbidity and mortality among HIV-exposed Indian infants. BMC Infect Dis 2011; 11:193.
27. Bertolli J, Hsu H-W, Sukalac T, Williamson J, Peters V, Frederick T, et al. Hospitalization trends among children and youths with perinatal human immunodeficiency virus infection, 1990–2002. Pediatr Infect Dis J 2006; 25:628–633.
28. Judd A, Doerholt K, Tookey PA, Sharland M, Riordan A, Menson E, et al. Morbidity, mortality, and response to treatment by children in the United Kingdom and Ireland with perinatally acquired HIV infection during 1996–2006: planning for teenage and adult care. Clin Infect Dis 2007; 45:918–924.
29. Crum-Cianflone NF, Grandits G, Echols S, Ganesan A, Landrum M, Weintrob A, et al. Infectious Disease Clinical Research Program. Trends and causes of hospitalizations among HIV-infected persons during the late HAART era: what is the impact of CD4 counts and HAART use?. J Acquir Immune Defic Syndr: JAIDS 2010; 54:248–257.
30. Buchacz K, Baker RK, Moorman AC, Richardson JT, Wood KC, Holmberg SD, et al. Rates of hospitalizations and associated diagnoses in a large multisite cohort of HIV patients in the United States. AIDS 2008; 22:1345–1354.1310.1097/QAD.1340b1013e328304b328338b.
31. Sansom S, Anderson JE, Farnham PG, Dominguez K, Soorapanth S, Clark J, et al. Updated estimates of healthcare utilization and costs among perinatally HIV-infected children. JAIDS: J Acquir Immune Defic Syndr 2006; 41:521–526.
32. Harling G, Wood R. The evolving cost of HIV in South Africa: changes in healthcare cost with duration on antiretroviral therapy for public sector patients. JAIDS: J Acquir Immune Defic Syndr 2007; 45:348–354.
33. Zhou F, Kominski G, Qian H-Z, Wang J, Duan S, Guo Z, et al. Expenditures for the care of HIV-infected patients in rural areas in China's antiretroviral therapy programs. BMC Med 2011; 9:6.
34. Smith de Cherif T, Schoeman J, Cleary S, Meintjes G, Rebe K, Maartens G. Early severe morbidity and resource utilization in South African adults on antiretroviral therapy. BMC Infec Dis 2009; 9:205.
35. Violari A, Cotton M, Gibb D. Early antiretroviral therapy and mortality among HIV-1 infected infants. N Engl J Med 2008; 359:2233–2244.
36. WHO. Report of the WHO Technical Reference Group Paediatric HIV/ART Care Guildeine Group Meeting; 2008. http://www.who.int/hiv/pub/paediatric/WHO_Paediatric_ART_guideline_rev_mreport_2008.pdf. [Accessed 16 May 2011].
37. Ciaranello A, Park J-E, Ramirez-Avila L, Freedberg K, Walensky R, Leroy V. Early infant HIV-1 diagnosis programs in resource-limited settings: opportunities for improved outcomes and more cost-effective interventions. BMC Med 2011; 9:59.
38. World Health Organization. Global HIV/AIDS response: epidemic update and health sector progress towards Universal Access. Progress Report 2011. http://www.who.int/hiv/pub/progress_report2011/en/index.html. [Accessed 16 May 2011].
Keywords:

antiretroviral therapy; children; cost; HIV; hospitalization

Supplemental Digital Content

Back to Top | Article Outline
© 2012 Lippincott Williams & Wilkins, Inc.