treat patients with diabetes, not their numbers (esp the elderly!!)

Studies have shown that older patients with diabetes and hypertension (and hyperlipidemia, to a lesser degree) have increased likelihood of cognitive impairment as well as brain atrophy (also a predictor of subsequent cognitive decline). As part of the ACCORD trial (RCT of 10K diabetic pts with prevalent CAD or lots of risk factors, treated with intensive vs less-intensive glucose control, as well as greater or lesser lipid and blood pressure control), the Memory in Diabetes (MIND) substudy also assessed cognitive and MRI outcome of total brain volume (TBV) (see htn tight control DM inc cognitive decline jama int med 2014 in dropbox, or doi:10.1001/jamainternmed.2013.13656). Cognitive assessment was for psychomotor function and speed of learning/working memory (Digit Symbol Substitution Test), verbal memory and executive function (Rey Auditory Verbal Learning Test and Stroop Color-Word Test) and Mini-Mental Status Exam. 2977 participants without baseline cognitive impairment and HgbA1C <7.5% were randomized to systolic BP<120 vs <140, or to fibrate vs placebo in those with LDL<100 on statin. cognition assessed at baseline, 20 and 40 months. a subset of 503 had brain MRI to assess change in total brain volume (TBV). results:

    --baseline mean A1C=8.3%. mean age 62. mean duration of DM 10 yrs.

    --at 40 months, no diff in cognitive function in the intensive/less intensive BP-lowering or fibrate/placebo trial groups

    --at 40 months, TBV declined more in the intensive vs standard BP lowering groups. no diff in the fibrate group vs placebo

a more recent article in JAMA internal medicine assessed the number of ER visits for insulin-induced hypoglycemia and errors (IHE), using the National Electronic Injury Surveillance System-Cooperative Adverse Drug Event Surveillance project to assess adverse drug events in insulin-treated patients, and the National Health Interview Survey to assess the national household use of insulin (see dm ER visits hypoglycemia JAMA int med 2014 in dropbox, or doi:10.1001/jamainternmed.2014.136). based on 8100 adverse event cases from 2007 through 2011, there were

    --an estimated 97,648 ER visits for IHE's annually (based on 9.2% of all ER visits, from the surveillance data). approx 1/3 resulted in in hospitalization. 

    --severe neurologic sequelae in 60.6% (32% with altered mental status, 23% with loss of consciousness/seizure, 5% with fall or injury etc)

    -- >50% with blood glucose <50 mg/dl

    --insulin treated pts >=80yo more than twice as likely to go to the ER (and for this age group, 12.4% of total ER visits were for IHEs, with 34.9 ER visits per 1000 insulin-treated patients) and 5x rate of hospitalization, vs patients 45-64yo 

    --most common cause was reduced food intake and administration of wrong insulin

so, these 2 articles are similarly disturbing  -- there is such a push now to get the numbers correct that we are often missing the big picture. some of this focus on A1C or blood pressure is that it is really easy for outside agencies (health insurors, fed govt) to get this data and to use it as a means to show that these agencies are really interested in measuring and improving quality.  this is not to say that there should be no assessment of quality, but that focusing on numbers (instead of patients with numbers) may be subtly or not-so-subtly pushing us to be too aggressive (and, there are significant incentives for us to do well with the numbers/disincentives if we don't).  as mentioned in prior blogs and what makes sense clinically to most of us (i think) is that we need to treat the overall patient. especially in the elderly, the gain of really tight A1c or blood pressure is far outweighed by the potential downside.  i do have some very elderly patients who are very stable with phenomenal A1c's or blood pressure readings, and with some angst (both the patient and i like seeing good numbers), i do pretty regularly decrease their BP or diabetic meds (and even try to get them off insulin, so long as their blood sugars don't soar to >250 or so, or become symptomatic), but i must admit that it is still a bit jarring for the A1c to increase from 6.2 to 8, or BP from 115/70 to 140-50/80.  these papers above reinforce the potential harm in insulin in elderly as well as potential worsening of cognitive function with too-tight BP control.