Abstract
Mobile health (mHealth) interventions provide a low‐cost, scalable approach to supporting parents with infant feeding advice with the potential to reduce health care visits and associated costs for infant feeding support. This Australian study examined the impact of the Growing healthy (GH) app on health service utilisation and out‐of‐pocket costs for families in the first 9 months of their infants life. A quasi‐experimental study with a comparison group was conducted in 2015–2016 with an mHealth intervention group (GH app, n = 301) and a nonrandomized usual care group (n = 344). The GH app aimed to support parents of young infants with healthy infant feeding behaviours from birth to 9 months of age. App‐generated notifications directed parents to age‐and feeding‐specific content within the app. Both groups completed surveys at baseline when infants were less than 3 months old (T1), at 6 months (T2) and 9 months (T3) of age. At T3, participants reported health services used and any out‐of‐pocket costs for advice on infant feeding, growth or activity. App users had lower odds (odds ratio: 0.38 95% confidence interval: 0.25, 0.59) of using one or more services and had lower number of visits to a general practitioner (1.0 vs. 1.5 visits, p = 0.003) and paediatrician (0.3 vs. 0.4 visits, p = 0.049) compared to the usual care group. There was no difference in out‐of‐pocket costs between groups. Provision of an evidenced‐based infant feeding app may provide substantial savings to the health system and potentially to parents through fewer primary health care and paediatrician visits.
Keywords: economics medical, health services, infancy, mHealth, obesity prevention, parents
Universal access to an evidence‐based infant feeding app such as Growing healthy may provide substantial savings to the health system and potentially to parents through fewer primary health care visits.
Key messages
Parents of infants are frequent users of primary health services for reassurance and general information about normal infant behaviour especially around feeding and growth.
Access to an evidence‐based infant feeding app may provide substantial savings to the health system and potentially to parents through fewer primary health care visits.
Universal access to an app such as Growing healthy may reduce inequities between parents who can afford out‐of‐pocket costs and those who cannot.
1. INTRODUCTION
The use of mobile delivered health interventions (mHealth) in maternal and child health care is increasing worldwide, with the potential for high reach, low cost and reduced burden on health services (Chen et al., 2018). A systematic review of 245 studies of mHealth interventions in the maternal and child health field revealed that around half of the interventions utilized a mobile phone app and just over a third included a component of health education or promotion (Chen et al., 2018). There is some evidence that mHealth interventions can improve health care utilization in low‐income countries (Sondaal et al., 2016) and result in improvements in some outcomes such as breastfeeding (Chen et al., 2018; Lee et al., 2016) and vaccination rates (Sondaal et al., 2016) in low resource settings. Much less is known, however, about the impact of these interventions on health care usage and costs in high‐income countries, where there is generally better access to essential antenatal and postnatal services.
Evidence suggests that in high‐income countries, mothers and their infants use more health care services in the early months of life, compared to any other time in childhood (Ou et al., 2017). A study in Melbourne, Australia, found that between 1996 and 1999, infants in their first 12 months of life had on average 35.7 visits across all health services, with 31% being general practitioner (GP) visits (Goldfeld et al., 2003). In line with this, a study using data from the Longitudinal Study of Australian Children found that in the first 6 months of life, infants would visit a GP on average 7.7 times (Ou et al., 2017). A recent study (Hayes et al., 2019) of health care utilization amongst 350 children under 5 years of age from a disadvantaged area (measured by area level indicators) of Sydney, Australia, found that health care consultations and associated healthcare funder costs were higher in the first 2 years of life (mean of 12 visits per year, mean cost of $1400 AUD per child) compared to the next 3 years (mean of eight visits per year, mean cost of $900 AUD per child). In this study 86% of consultations were for primary care, with single parent and low‐income households being more frequent users of health care, particularly primary care. This mirrors international data (van der Hulst et al., 2022) from the Netherlands showing that neighbourhood deprivation and low income were consistently and independently related to higher primary healthcare costs in children aged 0–3 years. It has been hypothesized (Hayes et al., 2019) that these families may have less supports available to them and thus seek reassurance from primary health care services if concerns arise.
This increased use of health services in early life, particularly amongst more disadvantaged families, may be needed and beneficial. Increased use may, however, also contribute to additional financial burden via ‘out‐of‐pocket’ costs for consultations. Out‐of‐pocket costs are goods and services that are not covered by a health insurance plan (either a publicly provided, universal health insurance scheme such as Medicare in Australia, or a private health insurance plan), which require a service fee. Out‐of‐pocket costs include copayments for partially covered goods and services, and costs for noncovered health goods and services (Australian Institute of Health and Welfare, 2023). In 2020–2021, 15% of total funding for health expenditure in Australia was paid by individuals in the form of out‐of‐pocket costs at an average cost of $1293 per person (Australian Institute of Health and Welfare, 2022). This is relatively high compared to other high‐income countries such as the United States (11%), France (9%), Germany (12%) and New Zealand (12%) (World Bank, 2022). High out‐of‐pocket costs have the potential to reduce access to health care services for socioeconomically disadvantaged parents, who often have higher health care needs (Jeyendra et al., 2013).
Interestingly, just under half of the visits to GPs and 70% of the visits to maternal and child health nurses in Australia are unrelated to illness, suggesting that these services may be accessed for parental advice and support around infant care, including sleep and settling, infant feeding and growth (Goldfeld et al., 2003). Around one‐quarter of mothers reported infant feeding problems in the first 6 months of their infant's life (Cook et al., 2019) and this is associated with health service usage and costs including for maternal and child health nurses, GPs and paediatricians (Le et al., 2016). In addition, qualitative research with both GPs (Brodribb et al., 2013; Jeyendra et al., 2013) and mothers (Hartley et al., 2012) suggests that there are a number of barriers to the provision of well child care by GPs. Barriers include access and cost of care, the limited time available during consultations to discuss nonmedical issues, a lack of consistent guidelines and GPs' lack of knowledge of anticipatory guidance related to infant care.
This raises an important question of whether the use of online sources such as apps/websites could provide a low‐cost, high‐reach mechanism for providing parent advice and support around well baby care and whether this may reduce the burden on the health system and out‐of‐pocket costs for parents. Despite the increased use of mHealth interventions to support parents, we are unaware of any previous studies in high‐income countries that have examined the impact of an mHealth intervention for parents on health service utilization and out‐of‐pocket costs.
The primary aim of the Growing healthy (GH) study was to establish the feasibility of providing evidence based, best practice information and support for healthy infant feeding practices using an mHealth (app) approach. This economic substudy considered the impact of the GH mHealth intervention on health service usage and out‐of‐pocket costs for families in the first 9 months of their infants’ life.
2. METHODS
2.1. GH feasibility study
Full details of the GH study have been published elsewhere (Denney‐Wilson et al., 2015). In brief, the study was conducted in 2015–16 and utilized a quasi‐experimental design with a comparison group (Office for Health Improvement and Disparities, 2021) that included an mHealth intervention group and a concurrent nonrandomized usual care group. The intervention aimed to support parents of young infants with healthy infant feeding behaviours. Eligibility criteria included: expectant parents (>30 weeks of gestation) or parents with an infant <3 months old, ability to read and understand English, ownership of a mobile phone, ≥18 years old and living in Australia. Recruitment to the GH mHealth programme consisted of: (1) practitioner‐led recruitment through Maternal and Child Health nurses, midwives and nurses in general practice; (2) face‐to‐face recruitment by researchers; and (3) online recruitment (Laws et al., 2016). A concurrent comparison group (baby's first foods [BFFs]) was recruited via social media using online parenting forums, blogs and Facebook. The BFF group received usual care and were not provided access to the app. The name BFF was applied to this cohort as we recruited participants separately and wanted to distinguish them from the intervention group; they were the comparison group for analysis purposes.
The intervention provided participants with an app, website and online forum to provide best practice information and support about infant feeding for the first 9 months of life. Participants received three text messages/push notifications a week, one related to milk feeding (breast, formula or mixed feeding), one related to other aspects of infant care (e.g., solids, sleep) and one targeted to mothers’ wellbeing. Messages were tailored to baby's age and feeding method. Participants also had access to a private Facebook page. The comparator group continued with ‘usual’ methods of obtaining information and assistance with infant diet, growth feeding and physical activity. The findings of the main feasibility study have been published elsewhere (Laws et al., 2018; Russell et al., 2018) but in brief, the GH app was found to be a feasible and acceptable mode for delivering obesity prevention intervention to parents. However, app usage declined over time and there were no significant differences between groups in any of the target behaviours or growth trajectories.
2.2. Measures
Data were collected via an online survey completed by the primary carer upon recruitment to the study (baseline T1), when the infant reached 6 months (T2) and 9 months of age (T3). The baseline survey collected information on family, maternal and child characteristics (e.g., infant's age and sex, whether first born, infant birth weight, Aboriginal status of mother and child, maternal age, maternal country of birth, maternal education, marital status, household income and health care card status).
At T3 (when the infant was 9 months), carers were asked whether they sought advice about their infant's growth, diet, feeding or activity. If yes, participants were asked about which health services they had sought this advice from including: GPs, lactation consultants, additional Maternal and Child Health Nurse appointments (outside of standard schedule of visits), Maternal Child Health Nurse helplines, paediatricians, home visiting/outreach nurses, chiropractors/osteopaths/naturopaths, parenting centres (day visit), parenting centres (night visit) and other health professionals (e.g., dentists, physiotherapy, speech pathology and specialists). Information was also sought regarding the number of times a service was utilized and the out‐of‐pocket cost (if applicable) for a single visit to that service. Costs were valued from the participant perspective in Australian dollars in 2016 (A$1 = US$0.759 at 1 January 2016).
2.3. Statistical analysis
All analysis was conducted using STATA 17. Descriptive statistics were conducted to summarize cohort characteristics. Continuous or categorical variables are presented as means and SD or percentages. χ 2 test and independent t test assessed the between‐group differences in categorical and continuous variables, respectively.
Multivariable logistic regression models with adjustment for potential covariates were conducted to compare percentages of health service use between intervention (GH) and comparison group (BFF). For percentages ≤10%, multivariable logistic regression with firth estimator was conducted using the ‘firthlogit’ command. The number of different service types used, total number of visits to all service types and the number of visits to each service type between GH and BFF were compared using Mann–Whitney U test and Zero‐inflated negative binomial regression model to account for non‐normal data distribution and excess zeros (no service use). Zero‐inflated model has two parts that use a negative binomial model to predict the count process (number of service use) and a logit model to predict zeros (no service use), respectively.
As some health services utilized in the first year of life are provided at no out‐of‐pocket cost to families (including government‐funded maternal child health visits and government‐ or third‐party payer‐funded Maternal Child Health nurse helplines) these health services were excluded from the out‐of‐pocket cost analysis. Percentages of participants with out‐of‐pocket costs for all relevant services were compared by multivariable logistic regression. Total out‐of‐pocket expenditure per infant was calculated by multiplying the number of visits with the reported out‐of‐pocket cost per visit. Due to the high variation in out‐of‐pocket costs and the potential implausibility of some reported values, mean plausible out‐of‐pocket costs per visit for health services were sourced from the grey literature (Comcare, 2023; Royal Australian College of General Practitioners, 2022) and adjusted using the Health Price Index (Australian Institute of Health and Welfare, 2023) (see Supporting Information S1: File 1). Those reporting ≥3 times the mean out‐of‐pocket cost were considered as implausible values and were coded as missing. Independent t test was used to test for differences between the GH and BFF groups for mean total out‐of‐pocket costs for all relevant services and each individual service. We adjusted for child age, whether first born and maternal age as covariates (potential confounders), to account for between‐group differences at study baseline. In addition, as previous research has shown differences in health care service access by sex (Simons et al., 2023), cultural groups (Khatri & Assefa, 2022) and education level (Butler et al., 2024), we also adjusted for maternal country of birth (as a proxy for cultural background) and maternal education in multivariable adjusted models. Analyses were conducted using data from all participants in the economic substudy.
2.4. Ethics statement
Ethics approval was provided by Deakin University 2014‐093 and University of Technology Sydney 2014000123.
3. RESULTS
3.1. Sample characteristics
In total, 909 parents undertook screening to participate in the study with 645 parents (GH = 301, BFF = 344) meeting the eligibility criteria (Laws et al., 2018). All participants were mothers. Of the original study cohort (n = 645), 202 (67.1%) and 277 (80.5%) of the GH and BFF groups completed the service use and economic component of the survey and were included in the current analyses. Cohort characteristics of participants by GH and BFF are presented in Table 1. Infants in the GH group were 1‐week younger and were more likely to be first born than those in the BFF group (p < 0.05). Mothers in the GH group were 1 year younger than mothers in the BFF group (p < 0.05) (Table 1). Other cohort characteristics including infant sex, birth weight, Aboriginal status, maternal marital status, education, whether has health care card and household income were similar between GH and BFF.
Table 1.
Child, maternal and household characteristics of economic substudy participants (n = 479), collected at baseline.
| GH | BFF | ||
|---|---|---|---|
| n = 202 | n = 277 | p | |
| Infant characteristics | |||
| Age (weeks), M (SD) | 6.9 (3.8) | 7.9 (3.8) | 0.004 |
| Sex (male, %) | 51.0 | 50.2 | 0.86 |
| Birth weight (kg), M (SD) | 3.5 (0.6) | 3.5 (0.7) | 0.38 |
| Birth order (first born, %) | 57.9 | 40.1 | <0.001 |
| Aboriginal or Torres Strait Islander status (%) | 2.0 | 1.8 | 0.89 |
| Maternal characteristics | |||
| Mother's age (years), M (SD) | 30.4 (4.6) | 31.3 (4.1) | 0.02 |
| Marital status (married, %) | 96.0 | 98.2 | 0.15 |
| Education level (completed university, %) | 51.5 | 53.8 | 0.63 |
| Has Health Care Card (Yes, %) | 14.9 | 13.0 | 0.56 |
| Country of birth (born in Australia, %) | 85.2 | 89.2 | 0.19 |
| Household characteristics | |||
| Household income (≥$100 K per annum, %) | 29.7 | 35.9 | 0.19 |
Note: Health Care Card is provided to individuals receiving various government welfare payments and enables recipient to receive concessions on the cost of some prescription medicines, medical services and other government concessions. Maternal age (n = 200 GH; n = 272 BFF); education (n = 194 GH; n = 277 BFF); household income (n = 175 GH; N = 234 BFF).
Abbreviations: BFF, baby's first food group; GH, Growing healthy.
3.2. Health service usage
The results for health service usage were first reported by likelihood of accessing services between groups and likelihood of accessing service, followed by the number of visits to services. Of the 479 participants completing the health service usage survey, 326 (68%) reported seeking advice from ≥1 health services about their infant's growth, diet, feeding or activity and 153 (32%) did not seek advice from any health professional (n = 84 GH group participants; 69 BFF group participants). GH participants had lower odds of using one or more health service than BFF participants in crude (odds ratio [OR]: 0.47, 95% confidence interval [CI]: 0.31, 0.69) and adjusted (OR: 0.38, 95% CI: 0.25, 0.59) models, respectively (Table 2). The health service that the highest proportion of participants in both groups reported seeking advice from was GPs (35% of GH participants and 48% of BFF participants). This was followed by extra Maternal and Child Health Nurse visits, lactation consultants, Maternal and Child Health Nurse helplines and paediatricians (Table 2). With adjustment for covariates, parents in the GH group had lower odds of seeking advice from a GP (OR: 0.51, 95% CI: 0.04, 0.75), Maternal Child Health Nurse helpline (OR: 0.60, 95% CI: 0.36, 0.99) or Paediatrician (OR: 0.59, 95% CI: 0.35, 0.99) than parents in the BFF group in adjusted models, respectively. No evidence of between‐group difference was found for seeking advice for other health services in either crude nor adjusted models (Table 2). Among those who sought advice from ≥1 health services (n = 326), results from both crude and adjusted models showed that GH participants had higher odds of seeking advice from lactation consultants and extra appointments with Maternal Child Health Nurses than BFF participants with adjusted OR of 1.75 (95% CI: 1.08, 2.84) and 2.24 (95% CI: 1.37, 3.66), respectively. For other health service usage, no differences were found between GH and BFF.
Table 2.
Health service usage for infant growth, diet, feeding or activity between GH intervention and BFF comparison group.
| Outcome | GH (%) | BFF (%) | Unadjusted OR | Adjusted ORa |
|---|---|---|---|---|
| Total sample (n = 479): GH 202, BFF 277 | ||||
| ≥1 health service | 58 | 75 | 0.47 (0.31,0.69)* | 0.38 (0.25, 0.59)* |
| GP | 35 | 48 | 0.60 (0.41,0.86)* | 0.51 (0.04, 0.75)* |
| Lactation consultant | 30 | 29 | 1.07 (0.72, 1.58) | 1.05 (0.68, 1.60) |
| Extra Maternal and Child Health Nurse appointments | 38 | 32 | 1.32 (0.90, 1.93) | 1.18 (0.78, 1.77) |
| Maternal Health Nurse helpline or other telephone helpline | 16 | 21 | 0.75 (0.47, 1.21) | 0.60 (0.36, 0.99)* |
| Paediatrician | 14 | 21 | 0.59 (0.36, 0.97)* | 0.59 (0.35, 0.99)* |
| Nurse home visit or outreach | 12 | 14 | 0.80 (0.46, 1.37) | 0.82 (0.46, 1.43) |
| Chiropractor/osteopath/naturopath | 9 | 13 | 0.66 (0.37, 1.20) | 0.69 (0.37, 1.29) |
| Parenting centre—day visit | 6 | 5 | 1.30 (0.60, 2.78) | 1.14 (0.52, 2.52) |
| Parenting centre—night visit | 6 | 4 | 1.52 (0.67, 3.46) | 1.42 (0.60, 3.39) |
| Dietitian | 2 | 3 | 0.79 (0.27, 2.29) | 0.62 (0.19, 2.00) |
| Other | 10 | 10 | 0.98 (0.54, 1.79) | 0.81 (0.43, 1.52) |
| In those using ≥1 service (n = 326): GH 118, BFF 208 | ||||
| GP | 60 | 63 | 0.87 (0.55, 1.38) | 0.81 (0.50, 1.35) |
| Lactation consultant | 52 | 38 | 1.71 (1.08, 2.70)* | 1.75 (1.08, 2.84)* |
| Extra Maternal and Child Health Nurse appointments | 65 | 42 | 2.56 (1.60, 4.09)* | 2.24 (1.37, 3.66)* |
| Maternal Health Nurse helpline or other telephone helpline | 28 | 27 | 1.02 (0.62, 1.70) | 0.82 (0.48, 1.43) |
| Paediatrician | 24 | 28 | 0.79 (0.47, 1.32) | 0.77 (0.44, 1.34) |
| Nurse Home Visit or outreach | 20 | 19 | 1.07 (0.61, 1.89) | 1.18 (0.65, 2.15) |
| Chiropractor/osteopath | 15 | 17 | 0.87 (0.47, 1.60) | 0.94 (0.50, 1.80) |
| Parenting centre—day visit | 11 | 7 | 1.72 (0.79, 3.74) | 1.45 (0.65, 3.26) |
| Parenting centre—night visit | 10 | 5 | 2.02 (0.87, 4.64) | 1.87 (0.76, 4.57) |
| Dietitian | 4 | 4 | 1.02 (0.35, 2.98) | 0.69 (0.21, 2.29) |
| Other | 17 | 13 | 1.32 (0.71, 2.45) | 1.12 (0.58, 2.15) |
Abbreviations: BFF, baby's first food; GH, Growing healthy; GP, general practitioner; OR, odds ratio.
Adjusted odds ratio included covariates: infant sex, infant age, whether first born, maternal age, education and country of birth. Firth estimator logistic regression was used for health service with percentage <10%. % are rounded to whole numbers.
*Denotes statistical significance.
The total number of different services used (1.8 vs. 2.0 services, p = 0.03) and the number of visits to GP (1.0 vs. 1.5 visits, p = 0.003) and paediatrician (0.3 vs. 0.4 visits, p = 0.049) were lower in the GH than in the BFF group (Table 3). Total number of visits to any service or other individual service (lactation consultant, Maternal and Child health Nurse helpline, nurse home visit, chiropractor/osteopath, parenting centre, dietitian and other) did not differ by GH and BFF (Table 3). However, when focusing on the sample who had visited ≥1 service, the GH group had a higher number of visits to lactation consultants (1.2 vs. 0.9 visits, p = 0.02) and extra Maternal and Child Health Nurse visits (1.6 vs. 1.2 visits, p = 0.02) compared to the BFF group. No statistically significant between‐group differences in the number of service types used, total number of visits to any service, or number of visits to other individual service types were found.
Table 3.
Mean total number of services used, total number of visits to each service and number of visits to each service by GH and BFF.
| Total sample (n = 479) | In those using ≥1 service (n = 326) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| GH (n = 202) | BFF (n = 277) | GH (n = 118) | BFF (n = 208) | |||||||
| Mean | Median | Mean | Median | Mean | Median | Mean | Median | |||
| (SD) | (IQR) | (SD) | (IQR) | p | (SD) | (IQR) | (SD) | (IQR) | p | |
| Number of different service types used | 1.8 (2.1) | 1 (0, 3) | 2.0 (1.8) | 2 (1, 3) | 0.03 | 3.1 (1.8) | 3 (1, 4) | 2.6 (1.5) | 2 (1, 4) | 0.08 |
| Total number of visits to any service | 5.0 (7.6) | 1 (0, 7) | 5.6 (7.1) | 3 (0, 8) | 0.14 | 8.6 (8.1) | 6 (3, 11) | 7.4 (7.3) | 5 (3, 10) | 0.28 |
| Mean number of visits to each service type | ||||||||||
| GP | 1.0 (1.8) | 0 (0, 1) | 1.5 (2.2) | 0 (0, 2) | 0.003 | 1.7 (2.1) | 1 (0, 3) | 2.0 (2.3) | 1 (0, 3) | 0.29 |
| Lactation consultant | 0.7 (1.5) | 0 (0, 1) | 0.7 (1.7) | 0 (0, 1) | 0.68 | 1.2 (1.8) | 1 (0, 2) | 0.9 (1.9) | 1 (0, 2) | 0.02 |
| Extra MCHN visits | 0.9 (1.8) | 0 (0, 1) | 0.9 (1.9) | 0 (0, 1) | 0.28 | 1.6 (2.1) | 0 (1, 2) | 1.2 (2.1) | 0 (0, 2) | 0.02 |
| MCHN helpline | 0.5 (1.7) | 0 (0, 0) | 0.4 (1.1) | 0 (0, 0) | 0.73 | 0.9 (2.1) | 0 (0, 1) | 0.5 (1.2) | 0 (0, 1) | 0.29 |
| Paediatrician | 0.3 (0.8) | 0 (0, 0) | 0.4 (1.1) | 0 (0, 0) | 0.049 | 0.5 (1.0) | 0 (0, 0) | 0.6 (1.3) | 0 (0, 1) | 0.42 |
| Health professional home visit or outreach | 0.3 (1.2) | 0 (0, 0) | 0.4 (1.4) | 0 (0, 0) | 0.48 | 0.6 (1.5) | 0 (0, 0) | 0.5 (1.6) | 0 (0, 0) | 0.69 |
| Chiropractor/osteopath | 0.5 (2.1) | 0 (0, 0) | 0.7 (2.3) | 0 (0, 0) | 0.2 | 0.9 (2.7) | 0 (0, 0) | 0.9 (2.7) | 0 (0, 0) | 0.72 |
| Parenting centre—day visit | 0.1 (0.4) | 0 (0, 0) | 0.1 (0.5) | 0 (0, 0) | 0.53 | 0.2 (0.5) | 0 (0, 0) | 0.1 (0.6) | 0 (0, 0) | 0.19 |
| Parenting centre—night visit | 0.4 (1.8) | 0 (0, 0) | 0.2 (1.0) | 0 (0, 0) | 0.31 | 0.6 (2.4) | 0 (0, 0) | 0.2 (1.2) | 0 (0, 0) | 0.09 |
| Dietitian | 0.04 (0.3) | 0 (0, 0) | 0.04 (0.3) | 0 (0, 0) | 0.79 | 0.07 (0.4) | 0 (0, 0) | 0.05 (0.3) | 0 (0, 0) | 0.85 |
| Other | 0.2 (0.9) | 0 (0, 0) | 0.2 (0.8) | 0 (0, 0) | 0.57 | 0.4 (1.1) | 0 (0, 0) | 0.3 (1.0) | 0 (0, 0) | 0.15 |
Note: Values presented as mean and SD. Tested using Mann–Whitney U test.
Abbreviations: BFF, baby's first food; GH, Growing healthy; GP, general practitioner; IQR, interquartile range; MCHN, Maternal and Child Health Nurse.
Results from the zero‐inflated negative binomial model are shown in Table 4. Significant between‐group differences were observed for use of the total number of different services and the total number of visits to any service in both unadjusted and adjusted models. In the adjusted model, among those who reported health service use, the GH group sought advice from a higher number of different services than the BFF group (coefficient: log[number of service] = 0.12, 95% CI: −0.05, 0.29). Moreover, the GH group was more likely to have a higher total number of visits to any service than BFF (coefficient: log[total number of visits] = 0.07, [−0.17, 0.31]). However, the GH also had a higher likelihood (positive coefficients) of having no service use (1.35, 95% CI: 0.70, 2.00) and no visits 1.20, 95% CI: 0.58, 1.80 than BFF For individual health services, results from both unadjusted and adjusted models revealed that the GH group had lower numbers (negative coefficients) of visits to GPs and higher likelihood (positive coefficient) of not visiting GP (zero visit) than BFF. No statistically significant between‐group difference was found for the number of visits to other health professionals.
Table 4.
Zero inflated negative binomial model to compare total number of services used, total number of visits to each service and number of visits to each service by GH and BFF.
| Unadjusted (n = 479) | Adjusted (n = 471) | |||
|---|---|---|---|---|
| Coefficient (predicting nonzeros) | Coefficient (predicting zeros) | Coefficient (predicting nonzeros) | Coefficient (predicting zeros) | |
| Number of different service types used | 0.19 (0.01, 0.36)* | 1.24 (0.58, 1.91)* | 0.12 (−0.05, 0.29) | 1.35 (0.70, 2.00)* |
| Total number of visits to any service | 0.16 (−0.08, 0.39) | 1.01 (0.42, 1.61)* | 0.07 (−0.17, 0.31) | 1.20 (0.58, 1.80)* |
| Number of visits to each service type | ||||
| GP | −0.14 (−0.43, 0.14) | 0.54 (0.06, 1.02)* | −0.18 (−0.48, 0.12) | 0.71 (0.19, 1.23)* |
| Lactation consultant | 0.03 (−0.56, 0.62) | −12.7 (−37.8, 12.3) | 0.02 (−0.44, 0.49) | −0.26 (−2.05, 1.53) |
| Extra MCHN visits | −0.31 (−0.76, 0.14) | −1.67 (−7.25, 3.92) | – | – |
| MCHN helpline | 0.64 (0.07, 1.20)* | 0.52 (−0.17, 1.20) | 0.54 (−0.11, 1.19) | 0.79 (−0.07, 1.65) |
| Paediatrician | −0.19 (−0.79, 0.43) | 0.46 (−0.29, 1.20) | −0.23 (−0.80, 0.32) | 0.43 (−0.28, 1.14) |
| Nurse home visit or outreach | 0.11 (−0.61, 0.83) | 0.32 (−0.42, 1.07) | – | – |
| Chiropractor/osteopath | 0.05 (−0.41, 0.51) | 0.40 (−0.21, 1.01) | −0.10 (−0.56, 0.36) | 0.34 (−0.31, 0.98) |
| Parenting centre—day visit | −0.7 (−1.96, 0.56) | −0.79 (−2.20, 0.62) | – | – |
| Parenting centre—night visit | 0.34 (−0.19, 0.86) | −0.4 (−1.24, 0.44) | – | – |
| Dietitian | 0.44 (−1.21, 2.09) | 0.46 (−1.18, 2.11) | – | – |
| Other | −0.03 (−0.55, 0.49) | −0.20 (−0.87, 0.48) | – | – |
Note: Model adjusted for child age, sex, whether first born, maternal age, education and country of birth. Model estimates are not shown due to convergence issues. For coefficients for predicting nonzeros (count model): positive coefficients represent GH more likely to visit than BFF and negative coefficients represent GH less likely to visit than BFF. For coefficients for predicting zeros, positive coefficients represent GH more likely to have zero visits than BFF, negative coefficients represent GH less likely to have zero visits than BFF.
Abbreviations: BFF, baby's first food; GH, Growing healthy; GP, general practitioner; MCHN, Maternal and Child Health Nurse.
*Denotes statistical significance.
3.3. Out‐of‐pocket costs
A similar proportion (53%) of participants in the GH and BFF reported that they paid out‐of‐pocket costs for visiting any service (Table 5). For individual health service, the highest proportion of participants experiencing out‐of‐pocket cost was observed when visiting Chiropractor/Osteopath, followed by paediatrician, parenting centre—night visits, GP, lactation consultant, dietitian and health professional home visits. Mean total out‐of‐pocket costs among those who used health services was $219 and $218 in GH and BFF, respectively (Table 6). Visiting a Chiropractor/Osteopath incurred the highest mean/median out‐of‐pocket cost, followed by visiting parenting centre—night visits, other health services, lactation consultant, paediatrician, dietitian and GP (Table 6). There were no statistically significant between‐group differences in total out‐of‐pocket cost for all services or any individual service.
Table 5.
Percentage of participants who used service that experience out‐of‐pocket costs by GH and BFF.
| Sample size (GH/BFF) | GH, n (%) | BFFs, n (%) | |
|---|---|---|---|
| Overall (used any service) | 98/205 | 52 (53%) | 108 (53%) |
| GP | 71/132 | 20 (28%) | 35 (27%) |
| Lactation consultant | 62/81 | 16 (26%) | 15 (19%) |
| Paediatrician | 29/59 | 20 (69%) | 41 (69%) |
| Nurse home visit or outreach | 24/40 | 0 (0%)a | 3 (8%) |
| Chiropractor/osteopath | 18/35 | 18 (100%) | 35 (97%) |
| Parenting centre—day visit | 13/14 | 0 (0%)a | 0 (0%) |
| Parenting centre—night visit | 12/11 | 4 (33%) | 6 (55%) |
| Dietitian | 5/9 | 1 (20%) | 4 (44%) |
| Other | 21/24 | 9 (43%) | 12 (50%) |
Abbreviations: BFF, baby's first food; GH, Growing healthy; GP, general practitioner.
No out‐of‐pocket costs reported.
Table 6.
Comparison of out‐of‐pocket costs in AUD by GH and BFF in those attending ≥1 service.
| GH | BFFs | ||||
|---|---|---|---|---|---|
| Mean (SD) | Median (IQR) | Mean (SD) | Median (IQR) | p | |
| Overall | 219.3 (483.9) | 0 (0, 250) | 218.1 (428.0) | 0 (0, 0) | 0.55 |
| GP | 33.2 (78.9) | 0 (0, 25) | 34.0 (91.7) | 0 (0, 0) | 0.28 |
| Lactation consultant | 114.8 (337.2) | 2 (0, 95) | 94.2 (364.6) | 0 (0, 0) | 0.26 |
| Paediatrician | 210.9 (251.0) | 160 (0, 305) | 272.1 (349.7) | 180 (0, 300) | 0.67 |
| Chiropractor/osteopath | 316.4 (279.0) | 210 (120, 440) | 382.3 (361.6) | 255 (125, 575) | 0.82 |
| Parenting centre—night visit | 316.2 (664.0) | 0 (0, 177.5) | 312.5 (400.0) | 150 (0, 640) | 0.41 |
| Dietitian | 4 (8.9) | 0 (0, 0) | 85.6 (115.5) | 0 (0, 190) | 0.24 |
| Other | 204.5 (358.8) | 0 (0, 240) | 263.4 (447.2) | 0 (0, 385) | 0.60 |
Note: p tested using Mann–Whitney U test.
Abbreviations: BFFs, baby's first foods; GH, Growing healthy; GP, general practitioner; IQR, interquartile range.
4. DISCUSSION
This is the first study to our knowledge examining the impact of an mHealth infant feeding intervention on parents’ use of health services and out‐of‐pocket costs for these services in the first 9 months of their infant's life. We found that after adjusting for covariates, participants using the GH app were 64% less likely to seek advice about infant feeding, growth and activity from one or more services, and were 50%, 42% and 44% significantly less likely to seek advice from a GP, Maternal Child Health Nurse helpline and paediatricians, respectively, compared to BFF group. This was further reflected in fewer total number of services used and lower number of visits to GPs and paediatricians in the GH app intervention group compared to the comparison (BFF) group.
These results suggest that having access to an app to provide first‐line information and support on infant feeding reduced the likelihood of parents seeking support from health services, particularly GPs and paediatricians for infant feeding issues. Given the high number of visits to GPs in the first year of life (Hayes et al., 2019; Ou et al., 2017), especially for issues unrelated to acute illness (Goldfeld et al., 2003), mHealth interventions such as GH have the potential to significantly reduce the burden on the primary health care system and reduce health care funder costs.
Despite reduced service usage in the app group, there was no difference in out‐of‐pocket costs for services used amongst app users compared to the usual care group. This might reflect the fact that the difference in service usage between the app and comparison group over the timeframe of this study was only 0.5 and 0.1 difference in GP and paediatrician visits, respectively. It is possible that providing app support for a longer duration could result in greater differences in out‐of‐pocket costs over time. Future research should assess the impact of mHealth interventions on health care utilization and out‐of‐pocket costs over a longer period of time, ideally up to 2 years of age because of the higher health care utilization during this phase of a child's life (Hayes et al., 2019).
It is also important to highlight that there has been a consistent decline in bulk billing for GP visits in Australia. For example, from 2022 to 2023, the proportion of GP patients who were always or usually bulk billed decreased from 89.0% to 77.3%, whereas the proportion of patients who were never bulk billed increased from 4.2% to 10.5% (Australian Institute of Health and Welfare, 2023). If this trend continues the amount of out‐of‐pocket costs for families with young children will continue to increase. This is a particular concern for families experiencing more disadvantage, who have been shown to be higher users of primary care services (Hayes et al., 2019; van der Hulst et al., 2022). Thus, interventions such as GH that reduce primary care consultations have important economic benefits, particularly for families experiencing socioeconomic disadvantage. This is even more critical given the escalating cost of living pressures coupled with the reduced earning capacity associated with caring responsibilities (Meadows et al., 2024). This could limit families’ capacity to pay for out‐of‐pocket healthcare costs and may result in families not seeking support for infant care.
Interestingly, among those who sought advice, the app group had higher service use/visits, but this was largely driven by lactation consultants and extra maternal and child health visits, suggesting parents sought more specialist rather than generalist advice. The app encouraged users to seek advice from lactation consultants for breastfeeding related problems and maternal and child health nurses for concerns regarding feeding, growth or development. Thus, it is not surprising to see a higher number of visits to these health professionals amongst app users as compared to the comparison group. This may reflect app users having better awareness of this type of assistance, desire to maintain breast feeding and/or more complex feeding issues in this group, all of which are worthy of future research. The higher use of these more specialist services in the app group did not result in any significant differences in out‐of‐pocket costs between groups. This may be because maternal and child health visits in Australia are free and while there are out‐of‐pocket costs for lactation consultants, there was only a small 0.3 visit difference between the two groups.
This study has a number of strengths and limitations. Our study adds important new information about the impact of a mHealth intervention on health care utilization and out‐of‐pocket costs associated with seeking support for infant feeding, growth and activity in the first 9 months of life. We included a broad range of services, with opportunities for participants to report any ‘other’ service used. Although the information provided is self‐reported by parents and subject to recall error, it does include all services including private services without a publicly funded component that would not be captured in linked data. We used best available data to look at plausible ranges of costs reported by participants, however we acknowledge that the ideal methodology would be to use linked data that captures the total cost of all services (e.g., health care funder, third party payer and out‐of‐pocket costs). Further, as recruitment to app group occurred from birth up to 3 months of age, exposure to the app also varied between participants which may have impacted the results. The results of this study were based on data collected in 2015–2016 and may not reflect current health service usage or out‐of‐pocket costs. However, as the only study examining the impact of mhealth intervention on these outcomes, we believe the study makes an important contribution to the literature. We also acknowledge that this was a feasibility study and not a randomized controlled trial powered to detect these effects. There were some baseline differences between the app and comparison group; however, all analysis has been adjusted for key covariates. Despite this, the study cannot establish a causal link between app usage and health care utilization and further research is required using a randomized controlled trial design to ascertain this relationship.
5. CONCLUSION
Parents using the GH mhealth infant feeding app were less likely to use health services for advice on infant feeding, growth or activity in the first 9 months of their child's life compared to those receiving usual care. App users who did seek additional help were more likely to seek support from specialized but less costly services such as lactation consultants and Maternal and Child health Nurses than the usual care group who were more likely to seek advice from a GP and/or Paediatrician. There was no difference in out‐of‐pocket costs between the two groups.
AUTHOR CONTRIBUTIONS
Rachel A. Laws, Elizabeth Denney‐Wilson, Sarah Taki, Catherine G. Russell, Karen J. Campbell all contributed to the conceptualization of the study and development of the app content. Kok‐Leong Ong developed the programming behind the app and website, and measurement of programme analytics. Eloise Litterbach managed overall data collection, whereas Sharyn Lymer developed health economics measures. Sharyn Lymer, Vicki Brown and Miaobing Zheng contributed to the analysis plan, Miaobing Zheng undertook the analysis and lead the writing of the health economics component of the study (Methods/Results). Rachel A. Laws wrote the first draft of the Introduction and Discussion, and managed all paper revisions. All authors reviewed and contributed to drafts of the paper and approved the final manuscript.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.
Supporting information
Supporting Information
ACKNOWLEDGEMENTS
The authors thank the parents who participated in the trial and the participating practitioners for their time in recruiting participants, and their valuable insights throughout the trial. We thank Leva Azadi for her early work on the breastfeeding components, Kate Dullaghan for her editorial work on the app content and Professor Cathrine Fowler for her support and review of app content. Thanks also to Louisa Wilson, our research assistant. The research reported in this paper is a project of the Australian Primary Health Care Research Institute, which was supported by a grant from the Australian Government Department of Health and Ageing. The information and opinions contained in it do not necessarily reflect the views or policy of the Australian Primary Health Care Research Institute or the Australian Government Department of Health and Ageing. Open access publishing facilitated by Deakin University, as part of the Wiley ‐ Deakin University agreement via the Council of Australian University Librarians.
Laws, R. A. , Zheng, M. , Brown, V. , Lymer, S. , Campbell, K. J. , Russell, C. G. , Taki, S. , Litterbach, E. , Ong, K.‐L. , & Denney‐Wilson, E. (2024). The impacts of an mHealth intervention targeting parents on health service usage and out‐of‐pocket costs in the first 9 months of life: The Growing healthy app. Maternal & Child Nutrition, 20, e13662. 10.1111/mcn.13662
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
- Australian Institute of Health and Welfare . (2022). Health expenditure Australia 2020‐21. Retrieved from: Canberra. Retrieved from: https://www.aihw.gov.au/reports/health-welfare-expenditure/health-expenditure-australia-2020-21/contents/summary
- Australian Institute of Health and Welfare . (2023). Health expenditure Australia 2021‐22 . Retrieved from: https://www.aihw.gov.au/reports/health-welfare-expenditure/health-expenditure-australia-2021-22/data
- Brodribb, W. , Zadoroznyj, M. , & Dane, A. (2013). The views of mothers and GPs about postpartum care in Australian general practice. BMC Family Practice, 14, 139. 10.1186/1471-2296-14-139 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Butler, D. C. , Larkins, S. , Jorm, L. , & Korda, R. J. (2024). Does use of GP and specialist services vary across areas and according to individual socioeconomic position? A multilevel analysis using linked data in Australia. BMJ Open, 14(1), e074624. 10.1136/bmjopen-2023-074624 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen, H. , Chai, Y. , Dong, L. , Niu, W. , & Zhang, P. (2018). Effectiveness and appropriateness of mHealth interventions for maternal and child health: Systematic review. JMIR mHealth and uHealth, 6(1), e7. 10.2196/mhealth.8998 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Comcare . (2023). Rates for medical and allied health treatment. Averaged across Australia for comparable services. Retrieved from: https://www.comcare.gov.au/service-providers/medical-allied-health/treatment-rates
- Cook, F. , Mensah, F. , Bayer, J. K. , & Hiscock, H. (2019). Prevalence, comorbidity and factors associated with sleeping, crying and feeding problems at 1 month of age: A community‐based survey. Journal of Paediatrics and Child Health, 55(6), 644–651. 10.1111/jpc.14262 [DOI] [PubMed] [Google Scholar]
- Denney‐Wilson, E. , Laws, R. , Russell, C. G. , Ong, K. , Taki, S. , Elliot, R. , Azadi, L. , Lymer, S. , Taylor, R. , Lynch, J. , Crawford, D. , Ball, K. , Askew, D. , Litterbach, E. K. , & J Campbell, K. (2015). Preventing obesity in infants: The Growing healthy feasibility trial protocol. BMJ Open, 5(11), e009258. 10.1136/bmjopen-2015-009258 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Goldfeld, S. , Wright, M. , & Oberklaid, F. (2003). Parents, infants and health care: Utilization of health services in the first 12 months of life. Journal of Paediatrics and Child Health, 39(4), 249–253. [DOI] [PubMed] [Google Scholar]
- Hartley, S. , Sutherland, G. , Brown, S. , & Yelland, J. (2012). ‘You're more likely to tell the GP if you're asked’: Women's views of care from general practitioners in the first postpartum year. Australian Journal of Primary Health, 18(4), 308–312. 10.1071/PY11111 [DOI] [PubMed] [Google Scholar]
- Hayes, A. J. , Brown, V. , Tan, E. J. , Chevalier, A. , D'Souza, M. , Rissel, C. , Baur, L. A. , Wen, L. M. , & Moodie, M. L. (2019). Patterns and costs of health‐care utilisation in Australian children: The first 5 years. Journal of Paediatrics and Child Health, 55(7), 802–808. 10.1111/jpc.14292 [DOI] [PubMed] [Google Scholar]
- van der Hulst, M. , Polinder, S. , Kok, R. , Prinzie, P. , de Groot, M. W. , Burdorf, A. , & Bertens, L. C. M. (2022). Socio‐economic determinants of healthcare costs in early life: a register‐based study in the Netherlands. International Journal for Equity in Health, 21(1), 5. 10.1186/s12939-021-01589-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jeyendra, A. , Rajadurai, J. , Chanmugam, J. , Trieu, A. , Nair, S. , Baskaran, R. , & Schmied, V. (2013). Australian general practitioners' perspectives on their role in well‐child health care. BMC Family Practice, 14, 2. 10.1186/1471-2296-14-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Khatri, R. B. , & Assefa, Y. (2022). Access to health services among culturally and linguistically diverse populations in the Australian universal health care system: Issues and challenges. BMC Public Health, 22(1), 880. 10.1186/s12889-022-13256-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- Laws, R. , Litterbach, E. , Denney‐Wilson, E. , Russell, C. G. , Taki, S. , Ong, K. , & Campbell, K. (2016). A comparison of recruitment methods for a mobile health intervention targeting mothers: Lessons from the Growing healthy program. Journal of Medical Internet Research, 2016, e248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Laws, R. A. , Denney‐Wilson, E. A. , Taki, S. , Russell, C. G. , Zheng, M. , Litterbach, E. K. , Ong, K. L. , Lymer, S. J. , Elliott, R. , & Campbell, K. J. (2018). Key lessons and impact of the Growing healthy mhealth program on milk feeding, timing of introduction of solids, and infant growth: Quasi‐experimental study. JMIR mHealth and uHealth, 6(4), e78. 10.2196/mhealth.9040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Le, H. N. D. , Gold, L. , Mensah, F. K. , Cook, F. , Bayer, J. K. , & Hiscock, H. (2016). Health service use and costs for infant behaviour problems and maternal stress. Journal of Paediatrics and Child Health, 52(4), 402–409. 10.1111/jpc.13095 [DOI] [PubMed] [Google Scholar]
- Lee, S. H. , Nurmatov, U. B. , Nwaru, B. I. , Mukherjee, M. , Grant, L. , & Pagliari, C. (2016). Effectiveness of mHealthmhealth interventions for maternal, newborn and child health in low‐ and middle‐income countries: Systematic review and meta‐analysis. Journal of Global Health, 6(1):010401. 10.7189/jogh.06.010401 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Meadows, J. , Montano, M. , Alfar, A. J. K. , Başkan, Ö. Y. , De Brún, C. , Hill, J. , McClatchey, R. , Kallfa, N. , & Fernandes, G. S. (2024). The impact of the cost‐of‐living crisis on population health in the UK: Rapid evidence review. BMC Public Health, 24(1), 561. 10.1186/s12889-024-17940-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Office for Health Improvement and Disparities . (2021). Quasi‐experimental study: Comparative studies. Retrieved from: https://www.gov.uk/guidance/quasi-experimental-study-comparative-studies
- Ou, L. , Chen, J. , & Hillman, K. (2017). Socio‐demographic disparities in the utilisation of general practice services for Australian children–Results from a nationally representative longitudinal study. PLoS One, 12(4):e0176563. 10.1371/journal.pone.0176563 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Royal Australian College of General Practitioners . (2022). General Practice Health of the Nation 2022. [Google Scholar]
- Russell, C. G. , Denney‐Wilson, E. , Laws, R. A. , Abbott, G. , Zheng, M. , Lymer, S. J. , Taki, S. , Litterbach, E. K. V. , Ong, K. L. , & Campbell, K. J. (2018). Impact of the growing healthy mhealth program on maternal feeding practices, infant food preferences, and satiety responsiveness: Quasi‐experimental study. JMIR mHealth and uHealth, 6(4), e77. 10.2196/mhealth.9303 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Simons, K. , Bradfield, O. , Spittal, M. J. , & King, T. (2023). Age and gender patterns in health service utilisation: Age‐period‐cohort modelling of linked health service usage records. BMC Health Services Research, 23(1), 480. 10.1186/s12913-023-09456-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sondaal, S. F. V. , Browne, J. L. , Amoakoh‐Coleman, M. , Borgstein, A. , Miltenburg, A. S. , Verwijs, M. , & Klipstein‐Grobusch, K. (2016). Assessing the effect of mHealth interventions in improving maternal and neonatal care in low‐ and middle‐income countries: A systematic review. PLoS One, 11(5), e0154664. 10.1371/journal.pone.0154664 [DOI] [PMC free article] [PubMed] [Google Scholar]
- World Bank . (2022). Out of pocket expenditue (% of health care expenditure). Retrieved from: https://data.worldbank.org/indicator/SH.XPD.OOPC.CH.ZS
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.