Table 5 Illustrative examples of the reporting of sex/gender in P4P impact studies

LeBlanc et al. [36]

− Results: “Among patients with baseline A1C levels between 6.5% and 7%, female patients had greater odds than males of receiving at least 2 A1C tests per year. Female physicians for all subgroups of patients were more likely than their male counterparts to order at least 2 A1C tests for their patients” (p. 193).

− Discussion: “…our findings suggest that patients followed by female family physicians may have better follow up in diabetes care. This finding is concordant with other studies that found that female physicians prescribe more laboratory tests than males” (p. 195)

Lippi Bruni et al. [37]

− Methods: “Patient demographics include dummies for gender and age classes. Other patient characteristics such as insulin dependence and number of visits to a diabetic outpatient clinic (DOC) are expected to capture severity. We control for GP gender, age and type of practice” (p. 143).

− Results: “…the area where the practice is located contributes to the variability between physicians more than the (observed) individual characteristics of the GP himself and of his group of patients. [Regarding the probability of emergency hospitalisations…] as for physician characteristics, age and postgraduate qualifications are not significant, whereas gender is significant and with a positive sign” (p. 145).

Iezzi et al. [6]

− Results: “Individual characteristics of the GP display certain effects [on the risk of diabetes complications], albeit not in a systematic manner. For instance, gender and seniority are not significant and neither practice type nor rural practice location produce any effect” (p. 112).

Yuan et al. [38]

− Background: “The purpose of our study was to investigate how the degree of glycemic control in patients with type 2 diabetes associates with lifestyle interventions as well as sociodemographic factors and further examine the differences by gender. … In addition, we analyzed whether inequalities in health status and disease control existed between genders” (p. 2).

− Results: “The average age of the female patients was greater than that of the male patients… Females were less well educated overall in this study population… [and] having physical activities (150 min/weekly) is more associated with the degree of glycemic control in males (P=0.003) than in females (푃=0.052)” (p. 3).

− Discussion: “The results of this study are intriguing and show that there appear to be sex-based differences in the stage and severity of diabetes... The impact of this health inequality seems to be related to socioeconomic conditions” (p. 8).

− Conclusion: “Health inequality is associated with gender and socioeconomic status in Taiwan and is disease-specific” (p. 10).

Hsiesh et al. [39]

− Results: “Regarding other [patient-level] confounding factors, men, older patients, patients with more severe comorbidity and patients with higher baseline density of cancer care tended to have higher risk of all-cause mortality” (p. 5).

Pan et al. [40]

− Methods: “The independent variables consisted of… personal characteristics of the research patients, including gender, age, and monthly salary” (p. e58).

− Results: “Compared with female patients, the COCI score of male patients was lower by 0.010 (P<.05)… Male patients showed a higher [hazard ratio] of mortality of 1.75 (95% CI, 1.71-1.80) compared with female patients” (p.e59).

Crawley et al. [41]

− Methods: “logistic regression was performed adjusting for age and gender” (p. 105).

− Discussion: “Our findings are consistent with several UK studies have examined equity in quality of care after the introduction of QOF using area-based measures of socioeconomic status… There is increasing evidence that inequities in care between age, gender and ethnic groups have persisted after the introduction of this pay for performance programme in the UK… Policy-makers and purchasers of healthcare should ensure that all such programmes are monitored for possible negative impacts on healthcare equity” (p. 106).

Millet et al. [42]

− Methods: “Patient-level variables were age, sex, ethnicity, neighborhood socioeconomic status (SES), and duration of diabetes” (p. 405).

− Results: “Pay for performance was associated with a significantly greater improvement in diastolic blood pressure in men than in women, but this pattern was reversed for A1C” (p. 407).

− Conclusions: “Our findings represent a more complete picture of disparities in diabetes management than that derived from national contract data, which lack patient level information on variables such as age, sex, ethnicity, and socioeconomic status and may underestimate variations in care… Our findings suggest that policy makers and health care planners should consider the potential negative impacts of pay for performance incentives on health care disparities” (p. 408).