Zhu ZZ 1997.

Methods	Generation of allocation sequence: unstated. Blinding: not used. Sample size estimation: no information. Withdrawal/drop‐out: unstated.
Participants	Ethnic: Chinese; 95 patients (53 in herbal treatment group, M/F 27/26, mean age 52 years (40‐70), mean disease duration 2.8 years (0.1‐5 years); 42 in control group, M/F 21/21, mean age 50 years (40‐70), mean disease duration 3 years (0.2‐5 years)). Setting: inpatients and outpatients. Inclusion criteria: type 2 diabetes mellitus, diagnosed by WHO criteria (WHO 1985), classified as deficiency of qi‐yin and deficiency of yin plus excess of heat. Exclusion criteria: unstated.
Interventions	Experimental intervention: Sanhuang Jiangtang recipe (mixture of 9 herbs) decoction, one dosage daily, divided into two times orally; plus tablet of the same herbal medicine, 8 tablets orally, t.i.d., for 4‐6 months. 12 out of 53 patients in this group maintained their hypoglycaemic drugs. Control intervention: glipizide, 5‐10 mg daily, 2‐3 times orally, adjusted based on blood glucose levels, the maximal dosage reached 30 mg daily; for 4‐6 months. Patients of the two groups were instructed for diet control and exercise.
Outcomes	Symptoms, FBG, and FINS. No follow‐up was reported.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear

FBG: fasting blood glucose; PBG: postprandial blood glucose; FINS: fasting insulin; HbA1c: glycated haemoglobin levels; ISI: insulin sensitivity index.