Prescriptions abstracted

In total, 5155 prescriptions were abstracted. The most common prescriptions were for anti-infective agents, central nervous system , and cardiovascular medications (Table 2) and these three categories accounted for over 7 0% of the medications dispensed. Approximately 70% of prescriptions were dispensed to Ontarians aged 65 years and older and 30% were dispensed to other recipients (people on social assistance and those with high prescription drug costs in relation to their income) (Table 3). Almost 90% of prescriptions were prescribed in clinics or offices, about 87% were written, and physicians prescribed over 98% (1.8% were written by dentists and one prescription was written by a chiropodist).

TABLE 2 Types of medications prescribed

Pharmacologic- Prescriptions abstracted therapeutic classification (n [%])
8:00 Anti-infective agents 1399 (27.1)
10:00 Antineoplastic agents 25 (0.5)
12:00 Autonomic agents 81
20:00 Blood formation and coagulation 35 (0.7)
24:00 Cardiovascular drugs 965 (18.7)
28:00 Central nervous system drugs 1287 (25.0)
36:00 Diagnostic agents 14 (0.3)
40:00 Electrolytic, caloric and water balance 194 (3.8)
48:00 Cough preparations 37 (0.7)
52:00 Eye, ear, nose and throat preparations 123 (2.4)
56:00 Gastrointestinal drugs 373 (7.2)
68:00 Hormones and substitutes 425 (8.2)
84:00 Skin and mucous membrane preparations 68
88:00 Vitamins 29 (0.6)
92:00 Unclassified therapeutic agents 97
99:00 Special classes (dietary supplements) 3 (0.1)
All 5155

*Numbers in the Pharmacologic-therapeutic column are from the American Hospital Formulary Service Pharmacological-Therapeutic Classification

TABLE 3 Sources and types of prescriptions

Characteristic Number of prescriptions (n=5155) (%)
ODB program
Social assistance 1557 (30.2)
Over age 65 years 3598 (69.8)
Source of prescription
Clinic/office 4626 (89.7)
Hospital 502 (9.8)
House call 9 (0.2)
Missing/other 18 (0.3)
Reported by
Fax 41 (0.8)
Paper 4509 (87.4)
Telephone 605 (11.8)
Prescriber type
Physician 5057 (98.2)
Dentist 93 (1.8)
Chiropodist 1
Missing 4

ODB Ontario Drug Benefit

Coding and prescribing errors

Of the 5155 dispensed prescriptions, a total of 37 errors was found, yielding an overall error rate of 0.7% (95% CI 0.5% to 0.9%) (Table 4). About 94% of prescriptions were dispensed as written on the prescription (Table 5). The amount dispensed in 137 prescriptions (2.7%) was reduced because the ODB did not cover the entire duration prescribed. In another 78 prescriptions (1.5%), there was a change due to packaging of the medication such that the prescribed quantity involved an amount that required partially filled bottles or leaving partially opened boxes of medications.

TABLE 4 Prescribing errors on information sent to the Ontario Drug Benefit (ODB) drug claims database

Type of error Number of prescriptions (n=5155) (%)
Prescribing error
Incorrect physician 11 (0.2)
Incorrect instructions 9 (0.2)
Incorrect quantity of pills 6 (0.1)
Incorrect dose 5 (0.1)
Incorrect drug (Garamycin Ophthalmic ointment 1 (0.0)
was dispensed when Sodium Sulamyd was prescribed)
Incorrect form (aqueous solution dispensed 1 (0.0)
when nasal spray was prescribed)
Clerical* 4 (0.1)

*Clerical errors included: copy of ODB label stuck to wrong prescription; patient name and medical doctor’s name not on prescription; two with incorrect dates of dispensation. Gentamicin sulfate (Garamycin Ophthalmic, Schering, Pointe Claire); sulfacetamide sodium (Sodium Sulamyd, Schering, Pointe-Claire)

TABLE 5 Changes in prescriptions made by pharmacists

Type of change Number of prescriptions (n=5155) (%)
Prescription size
Delivered as prescribed 4843 (93.9)
Reduced because of plan eligibility 137 (2.7)
Reduced – Reason unknown* 55 (1.0)
Reduced – 7 Day period 4 (0.1)
Enlarged 19 (0.4)
Reduced – Concern for drug abuse 1
Changed drug to one covered by ODB 18 (0.3)
Changed due to packaging 78 (1.5)

*After being reduced, the first prescription was between eight and 60 days. ODB Ontario Drug Benefit

Using both regular logistic and ‘rare events’ logistic regression, none of the characteristics of pharmacies examined (location, owner affiliation, productivity) was associated with coding errors.