Citations and methodology

Sources behind every number.

Every statistic on the Qiri site is traceable to a peer-reviewed paper, a government dataset, or an international body. Where we publish our own modelling, the per-site projections, we list the inputs and assumptions. If a number doesn't add up, tell us.

Jump to: Australia · United States · Global · Regulatory · Methodology

Australia

Backs the four problem-section statistics on qiri.ai/au.

  1. Felisberto M, et al. Override rate of drug-drug interaction alerts in clinical decision support systems: a systematic review and meta-analysis. Health Informatics J. 2024;30(2). journals.sagepub.com/doi/10.1177/14604582241263242
  2. Johnston K, O'Reilly CL, Scholz B, Georgousopoulou EN, Mitchell I. Burnout and the challenges facing pharmacists during COVID-19: results of a national survey. Int J Clin Pharm. 2021;43(3):716–25. link.springer.com/article/10.1007/s11096-021-01268-5
  3. Lim R, Kalisch Ellett LM, Semple S, Roughead EE. The Extent of Medication-Related Hospital Admissions in Australia: A Review from 1988 to 2021. Drug Saf. 2022;45(3):249–57. link.springer.com/article/10.1007/s40264-021-01144-1
  4. Australian Bureau of Statistics. Regional Population, 2023–24 financial year. Cat. no. 3218.0. abs.gov.au/statistics/people/population/regional-population/latest-release

United States

Backs the four problem-section statistics on qiri.ai/us, plus the workforce figure cited on /1995.

  1. Felisberto M, et al. Override rate of drug-drug interaction alerts in clinical decision support systems: a systematic review and meta-analysis. Health Informatics J. 2024;30(2). journals.sagepub.com/doi/10.1177/14604582241263242
  2. Kisala JR, et al. Evaluation of the Current State of Burnout Among Clinical Pharmacists. J Am Coll Clin Pharm. 2025. Supporting only: the population is clinical pharmacists, so it does not back the community-pharmacy burnout figure on qiri.ai/us, which uses the Mayo Well-Being Index below. accpjournals.onlinelibrary.wiley.com/doi/10.1002/jac5.70139
  3. Watanabe JH, McInnis T, Hirsch JD. Cost of Prescription Drug-Related Morbidity and Mortality. Ann Pharmacother. 2018;52(9):829–37. journals.sagepub.com/doi/10.1177/1060028018765159
  4. Guadamuz JS, Alexander GC, Chaudhri T, Trotzky-Sirr R, Qato DM. Locations and characteristics of pharmacy deserts in the United States: a geospatial study. Health Affairs Scholar. 2024;2(4):qxae035. academic.oup.com/healthaffairsscholar/article/2/4/qxae035
  5. Mayo Clinic Well-Being Index. State of Well-Being report, 2022–2023. 2023 data, 79,022 assessments: pharmacy professionals 62%, nurses 52%, physicians 51%. mywellbeingindex.org/downloads/state-of-well-being-2022-2023-report

Global

Backs the four problem-section statistics on qiri.ai/global.

  1. Felisberto M, et al. Override rate of drug-drug interaction alerts in clinical decision support systems: a systematic review and meta-analysis. Health Informatics J. 2024;30(2). journals.sagepub.com/doi/10.1177/14604582241263242
  2. Dee J, Dhuhaibawi N, Hayden JC. A systematic review and pooled prevalence of burnout in pharmacists. Int J Clin Pharm. 2023;45(5):1027–1036. link.springer.com/article/10.1007/s11096-022-01520-6
  3. Donaldson LJ, Kelley ET, Dhingra-Kumar N, Kieny MP, Sheikh A. Medication Without Harm: WHO's Third Global Patient Safety Challenge. The Lancet. 2017;389(10080):1680–1681. thelancet.com/journals/lancet/article/PIIS0140-6736(17)31047-4
  4. World Health Organization. Access to medicines and health products. Geneva: WHO. who.int/teams/health-product-policy-and-standards/access-to-medicines-and-health-products

Regulatory frameworks

Backs the regulatory pathway on qiri.ai/au/console. Qiri Phase 1 is designed against these frameworks; final classification is confirmed per deployment, not claimed.

  1. US Food and Drug Administration. Clinical Decision Support Software — Final Guidance. September 2022. Defines the four Non-Device CDS criteria under FD&C Act §520(o)(1)(E): no medical-image/signal analysis; displays patient medical information; provides recommendations to a healthcare professional; enables the HCP to independently review the basis. Qiri Phase 1 is architected against all four. fda.gov/regulatory-information/search-fda-guidance-documents/clinical-decision-support-software
  2. Australian Therapeutic Goods Administration. Regulation of software-based medical devices. Technology-agnostic SaMD framework. Phase 1 CDS software with licensed-pharmacist oversight typically classifies at lower risk. tga.gov.au/products/medical-devices/software-based-medical-devices
  3. European Union. Regulation (EU) 2024/1689 — Artificial Intelligence Act. Adopted June 2024; staged into force 2026–2027. Risk-classification framework for AI in healthcare. Qiri's pharmacist-in-the-loop and reasoning trace are designed against high-risk obligations: human oversight, transparency, record-keeping. eur-lex.europa.eu/eli/reg/2024/1689/oj
  4. Pharmacy Board of Australia (AHPRA). Code of conduct and Guidelines on dispensing of medicines. Defines the professional and legal standard for pharmacist oversight of dispensing in Australia, including the pharmacist's independent responsibility for clinical appropriateness, counselling, and final dispensing decisions. Qiri's pharmacist-in-the-loop architecture is designed so the pharmacist remains the regulated decision-maker under these standards: Qiri provides reasoning, the pharmacist provides the professional judgment the Code requires. pharmacyboard.gov.au/codes-guidelines.aspx

Methodology: per-site projections

The per-site model behind the return-on-intelligence figures, worked separately for each market.

The return-on-intelligence figures on the locale pages come from a Qiri internal model built on public benchmarks, not from external literature. Australia and the United States are modelled separately: different script volumes, different loaded wage rates, different currency, and only Australia carries funded clinical services. The two headline dollar figures are therefore built differently and should not be read against each other. The model is instrumented in live pilots now, and these figures are updated as pilot readouts land.

Inputs and arithmetic:

Input
Value
Basis
Script volume
AU ~200 a trading day
US ~250 dispenses a day
AU: 331.0M subsidised and under-co-payment prescriptions (FY2023–24) across ~5,900 community pharmacies, roughly 55–60,000 a year per pharmacy, per the PBS Expenditure and Prescriptions Report, 2023–24. US: a representative 250 against a typical 300–500. There is no sourced US scripts-per-day average we are willing to publish, so the US model deliberately runs below the typical range.
Time recovered per script
45–60 seconds
The model's only time assumption, and the same in both markets. Held deliberately below observational estimates of routine verification time (3–5 minutes per script).
Hours recovered
AU 750–1,000 a year
US 938–1,250 a year
Scripts × 45–60 seconds × ~300 trading days. That is 2.5–3.3 hours a day in Australia and 3.1–4.2 in the US. The locale pages round both bands down, to 2–3 and 3–4 hours.
Pharmacist loaded hourly rate
AU A$60–90
US US$65–85
Base salary plus ~25–30% on-costs (superannuation or benefits, leave loading, employer payroll obligations). AU from the Fair Work Pharmacy Industry Award (MA000012) base plus on-costs; US from BLS Pharmacists, Occupational Employment Statistics. The per-site model uses each pharmacy's actual loaded rate at deployment.
Wage value recovered
AU A$45–90K
US US$61–106K
Hours recovered × the loaded rate, at both ends of each band.
Funded services capacity
AU only, ~A$30K a year at the cap
Australian only: recovered time can be redirected into funded clinical services, and MedsCheck alone is capped at 20 services per pharmacy per month under the Pharmacy Programs Administrator. The US model claims no equivalent, because US reimbursement for pharmacist clinical services varies by state and payer and we have not modelled it.
Headline value per site
AU A$75–90K
US US$60–100K
The two figures are built differently and are not comparable. AU combines the conservative half of the wage band with partial use of funded services capacity. US is wage value only, rounded down from A$61–106K.
Indemnity
100% of decisions documented
Inputs, sources cited, rules fired, model used, pharmacist action, outcome. We do not publish a claims-reduction percentage: the effect on premiums and claims is being assessed with pilots and insurers, and until then the audit trail itself is the claim.
Validation
Live pilots, now
Pilots instrument time recovered per script and alert precision per site. Figures on this page are updated as pilot readouts land.

Market sizing (investor materials): global pharmacy management software and services is US$116.5B in 2026, growing to US$236.3B by 2031 at 15.2% CAGR, per Mordor Intelligence (July 2026), corroborated by The Business Research Company and 360iResearch; conservative scopes size the same category at US$34–53B. Dispensing labour: US pharmacies employ 321,970 pharmacists (mean wage US$140,920) and 471,680 pharmacy technicians (mean US$46,620), a combined wage bill of roughly US$67B a year, per BLS OEWS May 2025, occupation 29-1051 and 29-2052.

Ask Qiri

Backs the general-chatbot statistics on qiri.ai/ask-qiri. These figures describe general-purpose chatbots and AI search assistants, not purpose-built clinical software.

  1. Grossman S, et al. Appropriateness of ChatGPT as a resource for medication-related questions. Br J Clin Pharmacol. 2024. Long Island University College of Pharmacy posed 39 real drug-information queries to the free version of ChatGPT; pharmacists judged the responses incomplete or incorrect in nearly three-quarters of cases, and some answers cited references that did not exist. Presented at the ASHP Midyear Clinical Meeting, December 2023. doi.org/10.1111/bcp.16212
  2. Andrikyan W, et al. Artificial intelligence-powered chatbots in search engines: a cross-sectional study on the quality and risks of drug information for patients. BMJ Qual Saf. 2025;34(2):100–109. An expert panel judged 66% of AI search-chatbot answers to patient medicines questions as potentially harmful, 22% as potentially causing severe harm or death; only 54% aligned with scientific consensus. doi.org/10.1136/bmjqs-2024-017476
Qiri

Have a number we should add?

If you have peer-reviewed data, regulator publications, or actuarial evidence we should be citing, share it. We update this page when better sources land.

Send us a source Book a demo