The timing of insulin administration to hospital inpatients is unsafe: patient self‐administration may make it safer

Abstract

It is routine practice for insulin to be administered by hospital staff to diabetic inpatients who would normally self-inject at home. The accuracy of this practice has not been clearly evaluated. Our audit aimed to evaluate the accuracy of insulin administration times, patient preference and the possibility of self-administration by hospitalised patients.Information on patient preferences, and the timing of meals and insulin administration, was collected for insulin-treated diabetic inpatients whose insulin was administered by hospital staff (Group 1), and for another group whose insulin was self-administered (Group 2).In Group 1 (175 meals, 25 patients), only 19% of insulin administration times were deemed accurate. In this Group, 15/21 patients wanted to self-inject in hospital. In Group 2 (10 patients, 73 meals), the timing of 57/73 (78%) insulin self-administrations was accurate, and the remaining errors were often not through patient fault.The timing of insulin administration to hospitalised patients needs to be improved. An option may be to allow selected patients to self-administer their insulin. Copyright

Evaluating the Effectiveness of Self-Administration of Medication (SAM) Schemes in the Hospital Setting: A Systematic Review of the Literature

“…Of the 43 interventions, 10 SAM schemes had a single-staged approach to patient self-administration [21] , [22] , [25] , [31] , [36] – [41] , two had two stages [24] , [42] , 12 had three stages [9] , [10] , [20] , [26] , [27] , [30] , [32] , [34] , [43] – [46] , four had four stages [19] , [47] – [49] , two had five stages [28] , [33] , one had six stages [35] , one had seven stages [50] and one paper described an intervention comprising of nine stages of increasing patient independence [51] . For the remaining ten studies, it was not clear how many stages were used [23] , [29] , [52] – [59] . Most studies (n = 40) had an educational element to their SAM scheme.…”

“…Most studies (n = 40) had an educational element to their SAM scheme. For the remaining three studies, there was not enough information to determine whether an educational element existed [23] , [52] , [57] . In the majority of papers (n = 20) the educator was a pharmacist [20] – [22] , [27] – [30] , [32] , [34] – [36] , [38] – [40] , [44] , [46] , [48] , [50] , [54] , [56] , [58] , in seven a nurse [9] , [19] , [25] , [41] , [47] , [51] , [59] and in six both a nurse and pharmacist [10] , [26] , [42] , [43] , [49] , [55] .…”

“…Most studies did not specify which drugs were self-administered (n = 23) or simply allowed all drugs to be self-administered (n = 3). Commonly described self-administered drugs included psychotropic medicines [38] , [45] , [50] , insulin [23] , [34] and regular oral medication [34] , [56] . Drugs commonly excluded from SAM schemes were parenteral medication [10] , [20] , [28] , [29] , [31] , [39] , [49] , drugs of dependence [10] , [19] , [20] , [36] , [51] , insulin [10] , [29] , [31] , as required (PRN) drugs [19] , [36] , [49] , antibiotics [36] , [49] , and controlled substances [49] , [58] .…”

“…The terminology used (e.g. medication error vs. compliance) varied between studies however the outcome measures were often identical; therefore all of these studies are discussed together.Common methods for measuring this outcome included pill counts (or a comparable alternative) conducted by staff (n = 13) [20] , [22] , [23] , [27] – [30] , [32] , [33] , [39] , [41] , [47] , [54] and patient self-reported compliance (n = 10) [10] , [21] , [24] , [26] , [31] , [40] , [41] , [49] , [52] , [59] . One study measured patient compliance/error through patient urine samples [25] and one study measured compliance/error through disguised observation [58] .…”

“…One study measured patient compliance/error through patient urine samples [25] and one study measured compliance/error through disguised observation [58] . Of the studies, nine reported greater compliance or fewer errors in SAM groups compared with control groups [20] , [21] , [23] , [24] , [27] , [30] , [32] , [33] , [52] ; in five studies this difference was statistically significant [20] , [23] , [27] , [30] , [32] . Two of these five studies gained the equal highest quality assessment score for those studies measuring patient compliance.…”