Our goal is to conduct an outpatient study testing the bionic pancreas in the insulin-only configuration at a set-point of 100 mg/dl in 10 adult (≥ 18 years of age) subjects with type 2 diabetes in a random-order crossover study versus usual care with daily injections or an insulin pump.

The study will consist of two 7 day study arms: one usual care, and one insulin-only bionic pancreas at a set point of 100 mg/dl. The co-primary outcomes will be the mean Dexcom CGM glucose level and time <54 mg/dl, both in the last five days of each arm because these will be predictive of outcomes in long term use. The first two days will be excluded to allow for the extended washout of long acting insulins.

Inclusion Criteria:

Type 2 diabetes group: Age ≥18 years and clinical type 2 diabetes managed with:a multiple daily injection insulin regimen that includes NPH insulin and a rapid-acting insulin (insulin lispro, insulin aspart or insulin glulisine), an insulin pump filled with a rapid acting insulin, or a multiple daily injection insulin regimen that includes Lantus or Levemir and a rapid acting insulin.

Hemoglobin A1c >7%

Prescription medication regimen stable for > 1 month (except for medications that will not affect the safety of the study and are not expected to affect any outcome of the study, in the judgment of the principal investigator)

Live within a 60 minute drive-time radius of the central monitoring location

Willing to remain within a 120 minute drive-time radius of the central monitoring location throughout the study

Have someone over 18 years of age who lives with them, has access to where they sleep, is willing to be in the house when the subject is sleeping, and is willing to receive calls from the study staff and check the welfare of the study subject if telemetry shows a technical problem or severe biochemical hypoglycemia without subject response and the subject does not answer their telephone (up to two individuals can share this role, but they must be willing to carefully coordinate with each other and the subject so that one of them is clearly designated as having this responsibility at any given time)

Willing to wear one infusion set and one Dexcom CGM sensor and change sets frequently (an insulin infusion set every other day throughout the study) Have a mobile phone they are willing to keep with them and answer calls from study staff.

Exclusion Criteria:

Unable to provide informed consent (e.g. impaired cognition or judgment)

Unable to safely comply with study procedures and reporting requirements (e.g. impairment of vision or dexterity that prevents safe operation of the bionic pancreas, impaired memory, unable to speak and read English)

Current participation in another diabetes-related clinical trial that, in the judgment of the principal investigator, will compromise the results of this study or the safety of the subject

Pregnancy (positive urine HCG), breast feeding, plan to become pregnant in the immediate future, or sexually active without use of contraception Subjects must use acceptable contraception for the two weeks prior to the study, throughout the study and for the two weeks following the study. Acceptable contraception methods include: Oral contraceptive pill (OCP), Intrauterine Device (IUD, hormonal or copper), Male condoms, Female condoms, Diaphragm or cervical cap with spermicide, Contraceptive patch (such as OrthoEvra), Contraceptive implant (such as Implanon, Nexplanon), Vaginal ring (such as NuvaRing), Progestin shot (such as Depo-Provera), Male partner with a vasectomy proven to be effective by semen analysis

Need to go outside of the designated geographic boundaries during the study

Current alcohol abuse (intake averaging > 3 drinks daily in last 30 days), use of marijuana within 1 month of enrollment, or other substance abuse (use within the last 6 months of controlled substances other than marijuana without a prescription)

Unwilling or unable to refrain from drinking more than 2 drinks in an hour or more than 4 drinks in a day or use of marijuana during the trial

Unwilling or unable or to avoid use of drugs that may dull the sensorium, reduce sensitivity to symptoms of hypoglycemia, or hinder decision making during the period of participation in the study (use of beta blockers will be allowed as long as the dose is stable and the subject does not meet the criteria for hypoglycemia unawareness while taking that stable dose, but use of benzodiazepines or narcotics, even if by prescription, may be excluded according to the judgment of the principal investigator)

History of liver disease that is expected to interfere with the anti-hypoglycemia action of glucagon (e.g. liver failure or cirrhosis). Other liver disease (i.e. active hepatitis, steatosis, active biliary disease, any tumor of the liver, hemochromatosis, glycogen storage disease) may exclude the subject if it causes significant compromise to liver function or may do so in an unpredictable fashion.

Renal failure on dialysis

Personal history of cystic fibrosis, pancreatitis, pancreatic tumor, or any other pancreatic disease

Known history of coronary artery disease (CAD) that is symptomatic despite medical management including: unstable angina, angina that prevents moderate exercise (exercise of intensity up to 6 METS) despite medical management, myocardial infarction within the last 12 months of screening.

Known history of CAD that is not appropriately medically managed, e.g. not currently treated with ASA or other anti-platelet drug, a statin, and anti-hypertensives if indicated

Known history of CAD but participant is currently smoking tobacco

Abnormal EKG consistent with increased risk of malignant arrhythmia including, but not limited to, evidence of active ischemia, prior myocardial infarction, proximal LAD critical stenosis (Wellen's sign), prolonged QT interval (> 440 ms). Other EKG findings, including stable Q waves, are not grounds for exclusion as long as the participant is not exclude according to other criteria. A reassuring evaluation by a cardiologist after an abnormal EKG finding may allow participation.

Congestive heart failure with New York Heart Association (NYHA) Functional Classification III or IV

History of TIA or stroke in the last 12 months

Seizure disorder, history of any non-hypoglycemic seizure within the last two years, or ongoing treatment with anticonvulsants

History of hypoglycemic seizures (grand-mal) or coma in the last year

Untreated or inadequately treated mental illness (indicators would include symptoms such as psychosis, hallucinations, mania, and any psychiatric hospitalization in the last year), or treatment with anti-psychotic medications that are known to affect glucose regulation.

Electrically powered implants (e.g. cochlear implants, neurostimulators) that might be susceptible to RF interference

Unable to completely avoid acetaminophen for duration of study

Established history of allergy or severe reaction to adhesive or tape that must be used in the study

History of eating disorder within the last 2 years, such as anorexia, bulimia, or diabulemia or omission of insulin to manipulate weight

History of intentional, inappropriate administration of insulin leading to severe hypoglycemia requiring treatment

Use of oral anti-diabetic medications other than metformin

Lives in or frequents areas with poor Verizon wireless network coverage (which would prevent remote monitoring)

Any factors that, in the opinion of the principal investigator would interfere with the safe completion of the study

16 subjects were enrolled. 7 subjects participated. 5 subjects completed both arms of the study.

During each of the two 4-day study arms