Table 1.

Overview of studies employing the SGHS.

Reference	Sample Size; Locality	Key Findings
Browne et al. [44] a	1524; Australia	SGHS scores correlated with the full 72-item checklist at r = 0.94. Increasing SGHS scores predicted subjective wellbeing (r = −0.29) better than PGSI or addiction measures. Scale shown to be unidimensional and to have measurement invariance with respect to age and gender.
McLauchlan et al. [50] a	532; US	The psychometric performance of the SGHS remained equivalent when changing the scoring format from binary to Likert scale. No significant differences were found among correlations between the binary and Likert versions of the SGHS and measures of psychological distress, impulsivity, and wellbeing.
Murray-Boyle et al. [51] a	5551; Australia/New Zealand	The SGHS was strongly correlated with a range of measures, including the PGSI (r = 0.68), a latent gambling harm variable (r = 0.87), and a gambling harm scale only including unambiguously harmful consequences (r = 0.73). The findings lend support to the unidimensionality and reliability of the SGHS, particularly concerning the SGHS items capturing legitimate harmful consequences.
Murray-Boyle et al. [52] a	1742; Victoria, Australia	When examining five SGHS items criticised as being non-genuine harms, this study found that endorsing any of the five items predicted lower wellbeing and higher psychological distress. Each item individually predicted declines in health-related quality of life, and endorsement of additional harm items was associated with cumulative declines.
Acil Allen Consulting et al. [53] b	5000; Tasmania, Australia	Using the SGHS, 5531 years of life were lost due to gambling per annum, and this figure was similar when using the PGSI (5083 years of life lost). The mean number of harms increased along with PGSI categories: non-problem gamblers had a mean SGHS score of 0.057, 0.59 for low-risk gamblers, 2.164 for moderate-risk, and 5.565 for PGs.
Browne et al. [54] b	1174; Canada	Similar proportions of respondents scored 1+ on the PGSI (48.6%) compared to non-zero responses on the SGHS (41.9%). The key proximal and distal risk factors for gambling harm were trait impulsivity, early childhood gambling exposure, gambling fallacies, less use of safe gambling practices, and excessive gambling.
Browne and Rockloff [19] b	1524; Australia	This research demonstrated behavioural dependence as unidimensional and distinct from gambling harm. Nonetheless, harm mediated the relationship between behavioural dependence and wellbeing. Taken together, behavioural dependence and the SGHS predicted wellbeing better (10.2% explained variance) than each measure individually.
Dowling et al. [55] b	5000; Tasmania, Australia	The PGSI and SGHS, when considered separately, produced similar low-risk gambling guidelines and captured similar proportions of gamblers in the general population.
Hawker et al. [56] b	97; Tasmania, Australia	The proportion of gamblers who had experienced harm (1+ on the SGHS; 25.77%) was similar to those who had scored 1+ on the PGSI (23.71%).
Hing et al. [57] b	1174; Canada	This study used the SGHS as an outcome measure to develop nine safe gambling practices to best prevent GRH. Six practices were associated with reduced harm (e.g., I keep a household budget) and three were associated with increased harm (e.g., I have used cash advances on my credit card to gamble).
Hing et al. [58] b	92; Victoria, Australia	Gambling harms were negatively associated with saving behaviours related to money management (r = −0.34). No significant relationships existed between gambling harm and other aspects of financial literacy/money management (self-confidence, importance, knowledge, helping, and difficulties).
Jenkinson et al. [59] b	5076; Australia	The three most highly endorsed items from the SGHS were reduction of available spending money (24%), reduction of savings (22%), and regrets that made them feel sorry about their gambling (18%).
Newall et al. [60] b	789; UK	Custom sports bettors experienced a higher mean number of gambling harms compared to non-custom sports bettors (2.35 vs. 1.53). The SGHS was also highly correlated with the PGSI (rpb = 0.82).
Paterson et al. [61] b	5788; Australian Capital Territory, Australia	The 12-month prevalence of experiencing gambling harm was 9.6%. When comparing scores on the SGHS of 1+ (9.6%) to scores of 1+ on the PGSI (10.3%), no statistically significant differences were found. However, 8.7% of non-problem gamblers (PGSI) reported 1+ gambling harms on the SGHS.
Rockloff et al. [62] b	188; US	No significant interactions were found between PGSI status or gambling harm (SGHS) by free-spins influencing bet count.
Rockloff et al. [63] b	7626; Victoria, Australia	The prevalence of experiencing any GRH was 9.6%, with the most frequently endorsed harms being reductions in available spending money (5.1%), reduced savings (3.9%), and regrets about their gambling (3.4%). As the PGSI categories increased, so too did the proportion of having experienced harm. For non-problem gamblers, 4.3% had experienced harm alongside 29.2% of low-risk gamblers, 59.4% of moderate-risk gamblers, and 100% of problem gamblers.
Rodda et al. [64] b	104; Australia	Gamblers who busted (set a limit and broke it) experienced significantly more gambling harms than those who did not bust (4.26 vs. 0.86).
Russell et al. [65] b	2004; New South Wales, Australia	Almost half of the respondents (44.2%) had scored 1+ on the PGSI compared to 45.2% who nominated experiencing some GRH using the SGHS. Furthermore, both PGSI and SGHS scores were significantly associated with exposure to loot boxes.
Russell et al. [66] b	784; Victoria, Australia	Gambling harms were strongly related to the PGSI, with a positive relationship between mean SGHS scores and increasing PGSI categories.
Salonen et al. [67] b	2624; Finland	The prevalence of experiencing GRH was 11%, with emotional/psychological and financial domains of harm being notably impacted.
Woods et al. [68] b	5982; South Australia, Australia	Using the SGHS, the 12-month prevalence of experiencing any gambling harm was 19%, and it was higher among people in Greater Adelaide than the rest of the state.

^a evaluations of the SGHS; ^b applications of the SGHS.