The influence of socioeconomic status on growth hormone treatment in Australia

Abstract

Background: In Australia growth hormone (GH) is provided as a pharmaceutical benefit to any eligible child. The most common eligible conditions are GH deficiency (<10mU/L on 2 tests), Turner Syndrome (TS), and short stature (<1st centile) and slow growth (growth velocity <25th centile for sex and bone age) (SSSG). Socioeconomic status (SES) can be inferred from a patient's post code (Australian Bureau of Statistics; ABS). Demographic and treatment data are recorded in the OZGROW database. Aims: As GH is freely available to eligible children we would expect GH patients to be distributed evenly across SES strata. We were also interested to determine if SES was associated with other demographic or response measures. Methods: SES was determined from the ABS Postal Area Index of Relative Socio-economic Advantage and Disadvantage (2006). Patients (4318) ceasing GH between 1993 and 2013 were accessed from OZGROW. SES was interpreted in deciles: D1-lowest, D10-highest. Frequency distribution across SES deciles was hypothesised to be even and tested by Chi square. Tests were conducted on the whole population and within condition and sex. Between sex tests were also done. Regression analyses were used to test for associations between SES and starting height SDS (1stSDS), starting age (1stAge), 1st year response (1stdSDS), parental height (MPHSDS), last height SDS (LastSDS), total response (dSDS), and treatment duration (RxYs). Results: SES was positively associated with patient frequency (P=2.4×10-98). This was true for boys (P=1.4×10-53), girls (P=3.1×10-41), and each condition; GHD: P=2.2×10-17 (boys: P=6.0×10-9, girls: P=1.6×10-7); SSSG: P=3.2×10-49 (boys: P=6.1×10-34, girls: P=7.6×10-13); TS: P=4.5×10-17. In each case distributions were similar for boys and girls (P>0.77). It is expected that 10% of cases occur in each SES decile. In general, for each condition and sex, it was seen that D1 had 5% rising slowly to 7.5% in D4, jumping to 13% in D7, then levelling off to 15% (D10). SES was not associated with 1stdSDS. An increase in 1 SES decile was associated with (P<0.05) changes of -0.17 years 1stAge (TS only), +0.023 to +0.029 in 1stSDS, +0.023 to +0.025 for MPHSDS (except TS), +0.038 to +0.049 for LastSDS, +0.022 for dSDS (SSSG only), and +0.10 to +0.17 for RxYs (except GHD). Conclusions: Only 5% of GH patients come from the lowest SES decile while 15% come from the highest. The greatest increase in the frequency of patients occurred across D4 to D7. The disparity is concerning given the availability of GH in Australia. The disparity may reflect a city-country or a more general SES divide. That boys and girls were similarly affected suggests SES does not contribute to the well-known gender bias favouring boys in GH treatment. While 1stdSDS was not influenced by SES, total response measures favoured higher SES patients. Factors such as compliance and persistence may thus need to be investigated.L6135224012016-12-08 <br />