📄 medicine

The Digital Path Toward Perceived Medication Error Prevention From the Physician Perspective in Private Clinics in Eastern Indonesia: A Mixed Methods Study

This mixed-methods study of general practitioners in private clinics in Eastern Indonesia reveals that perceived medication error prevention is primarily driven by effective digital workflow alignment, which is significantly influenced by AI readiness, digital competency, EMR system quality, and peer influence, while remaining unaffected by capitation-based financing regulations.

Original authors: Debora Natalia, Waynhena Sarawa Nesia, Dewi Sri Surya Wuisan, Ferdi Antonio

Published 2026-07-07

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Debora Natalia, Waynhena Sarawa Nesia, Dewi Sri Surya Wuisan, Ferdi Antonio

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a doctor's office in a remote part of Eastern Indonesia as a busy kitchen. The doctor is the chef, and the patients are the orders coming in. In this kitchen, "medication errors" are like accidentally adding salt instead of sugar or giving a customer a dish they are allergic to. These mistakes can happen when the chef is rushing, the recipe book is messy, or the ingredients are hard to find.

This study asked: How does a digital recipe book (Electronic Medical Records, or EMR) help chefs (doctors) avoid these mistakes?

The researchers didn't just ask if the digital book existed; they asked if the chefs actually liked using it and if it fit naturally into their cooking rhythm. They studied 382 doctors in 41 private clinics.

Here is the breakdown of their findings, using simple analogies:

1. The "Perfect Fit" is the Secret Sauce

The biggest discovery was that the digital system only helps prevent mistakes if it fits the doctor's daily routine.

The Analogy: Imagine trying to cook a complex meal while wearing a pair of shoes that are two sizes too big. Even if the shoes are high-tech and expensive, they just slow you down and make you trip.
The Finding: Doctors felt the digital system prevented errors only when it felt like a natural extension of their hands—helping them write notes, check history, and prescribe medicine without stopping their flow. If the system felt clunky or forced, it didn't help, no matter how advanced it was.

2. The Chef's Skill Matters Most

The study found that the doctor's own digital skills were the most important factor in making the system work well.

The Analogy: You can give a master chef a high-tech, voice-activated knife, but if they don't know how to hold it or trust it, they won't use it. Conversely, a skilled chef can make a simple, old-fashioned knife work wonders.
The Finding: Doctors who felt confident using the computer, understanding the data, and fixing small glitches were the ones who felt the system was actually keeping patients safe. Their confidence made the digital tools "fit" their workflow.

3. The Quality of the "Kitchen Tools"

The quality of the system itself mattered. Was it fast? Did it crash? Was the information clear?

The Analogy: If the digital recipe book loads slowly every time you open it, or if the internet cuts out while you are ordering ingredients, the chef gets frustrated and stops using it.
The Finding: Even if a doctor was skilled, a slow or buggy system broke the "fit." When the system was reliable and easy to use, it helped prevent errors. When it was slow, doctors often ignored it, which increased the risk of mistakes.

4. The "AI Readiness" (Being Open to the Future)

The study looked at whether doctors were ready to use advanced features like Artificial Intelligence (AI) alerts.

The Analogy: This isn't about whether the kitchen has a robot chef yet. It's about whether the human chef is mentally prepared to trust a robot assistant that says, "Hey, that ingredient might be bad for this customer."
The Finding: Doctors who felt their clinic was ready for these advanced tools (good data, good policies, trust in the tech) were more likely to integrate the digital system smoothly. However, they only trusted the AI if the data behind it was accurate.

5. The "Peer Pressure" (But a Good Kind)

The study checked if colleagues influenced each other.

The Analogy: If everyone else in the kitchen is using the new digital tablet to check orders, you are more likely to pick one up and try it too.
The Finding: Seeing colleagues use the system successfully helped, but it was the least important factor compared to the doctor's own skills and the system's quality. It helped, but it wasn't the main driver.

6. The "Payment Rule" Didn't Change the Cooking

The researchers tested if the way doctors get paid (a system called Capitation, where they get a fixed amount per patient) changed how well the digital system worked.

The Analogy: They asked: "Does the fact that the restaurant owner pays the chef a flat fee for every customer change how well the chef uses the new digital knife?"
The Finding: No. The payment rule didn't make the digital system work better or worse. Whether the doctor was paid per patient or per visit, the success of the system still depended entirely on how well it fit their workflow and how skilled they were. The payment rule was just background noise, not a game-changer.

The Bottom Line

The paper concludes that simply handing a doctor a computer doesn't stop medication errors.

To make digital tools work in these clinics, you need three things working together:

Skilled Chefs: Doctors who are confident and trained to use the tech.
Good Tools: Systems that are fast, reliable, and easy to navigate.
A Good Fit: A system that slots perfectly into the doctor's daily routine without getting in the way.

If these three align, doctors feel safer and more confident that they are preventing mistakes. If they don't align, the technology is just a fancy distraction.

Technical Summary: The Digital Path Toward Perceived Medication Error Prevention in Eastern Indonesia

Problem Statement
In low-resource settings, particularly in Eastern Indonesia, primary care facilities face significant vulnerabilities regarding medication errors due to constrained capacities for reliable prescribing, monitoring, and follow-up. While Electronic Medical Records (EMRs) are increasingly mandated and adopted in private clinics, their mere presence does not guarantee improved safety. The core problem identified is the gap between EMR adoption and meaningful clinical integration. In environments with scattered populations and uneven digital infrastructure, EMRs often remain administrative tools rather than clinical instruments. Consequently, there is a lack of evidence regarding how socio-technical factors—specifically the alignment of digital systems with clinical routines (workflow fit)—influence physicians' perceptions of medication error prevention in private primary care settings.

Methodology
The study employs a sequential explanatory mixed-methods design involving 382 general practitioners (GPs) from 41 private clinics in remote Eastern Indonesia.

Quantitative Phase: A cross-sectional survey was analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) via SmartPLS 4. The conceptual model tested the relationships between four independent variables (AI readiness, digital competency, EMR system quality, and peer influence) and the dependent variable (perceived medication error prevention), mediated by digital workflow fit. Additionally, capitation-based financing regulation was tested as a moderator. The model utilized a bootstrapping method with 10,000 subsamples.
Qualitative Phase: Following the quantitative analysis, 15 GPs were purposively selected to answer open-ended questions. This phase aimed to interpret the quantitative findings, specifically exploring the mechanisms behind workflow integration, system barriers, and the role of financing regulations.
Validation: The measurement model was validated for reliability (Cronbach's alpha, composite reliability), convergent validity (AVE), and discriminant validity (HTMT). Predictive capability was assessed using Cross-Validated Predictive Ability Test (CVPAT).

Key Results
The structural model analysis yielded the following significant findings:

Determinants of Digital Workflow Fit: All four antecedent factors positively influenced digital workflow fit:
- Digital Competency: The strongest predictor ( $\beta = 0.399, p < 0.001$ ).
- AI Readiness: Significant positive association ( $\beta = 0.197, p = 0.008$ ).
- EMR System Quality: Significant positive association ( $\beta = 0.186, p = 0.002$ ).
- Peer Influence: Significant positive association ( $\beta = 0.123, p = 0.035$ ).
- Collectively, these factors explained 58.4% of the variance in digital workflow fit ( $R^2 = 0.584$ ).
Impact on Medication Error Prevention: Digital workflow fit demonstrated a strong, direct positive relationship with perceived medication error prevention ( $\beta = 0.747, p < 0.001$ ), explaining 57.9% of the variance in the outcome variable.
Moderation Effect: The hypothesis that capitation-based financing regulation moderates the relationship between digital workflow fit and perceived medication error prevention was rejected ( $\beta = -0.009, p = 0.388$ ). The data indicates that financing regulations do not significantly strengthen or weaken the link between workflow alignment and safety perceptions.
Qualitative Insights:
- Workflow Alignment: Physicians valued EMRs primarily when they facilitated rapid access to medication history and supported decision-making without disrupting consultation flow.
- Competency and Trust: Confidence in using EMR features was central to safety perceptions. Readiness for AI features was conditional on data accuracy and clinical trust.
- Infrastructure Barriers: Technical instability (slow systems, poor internet) was a critical barrier that undermined the utility of even high-quality systems.
- Regulatory Context: Capitation regulations were perceived as insufficient drivers for EMR adoption, often viewed as administrative burdens rather than safety enablers.

Key Contributions

Mediating Role of Workflow Fit: The study establishes digital workflow fit as a critical mediator, demonstrating that individual readiness, system quality, and social influence only translate to perceived safety when they successfully align with clinical routines.
Contextual Evidence from Low-Resource Settings: It provides empirical data from Eastern Indonesia, highlighting that in resource-constrained environments, the integration of technology is more predictive of safety perceptions than the adoption of technology or the presence of financing regulations.
Refutation of Regulatory Moderation: The study challenges the assumption that capitation-based financing directly moderates the effectiveness of digital workflows in preventing errors, suggesting that regulatory frameworks may act as contextual constraints rather than active moderators of the workflow-safety relationship.

Significance and Claims
The paper claims that the digital pathway to perceived medication error prevention in private primary care is fundamentally dependent on effective digital workflow alignment. The authors assert that:

Competency is Paramount: Improving doctors' digital skills is the most vital intervention for optimizing EMR utility.
Integration Over Implementation: Success depends less on the availability of EMRs and more on their seamless embedding into daily clinical tasks (prescribing, monitoring, follow-up).
Policy Implications: National mandates for EMR adoption must be accompanied by investments in training, system usability, and technical support. The study suggests that policymakers should focus on "meaningful use" through workflow redesign rather than relying solely on financing regulations to drive safety outcomes.
Modest Scope: The authors explicitly state that their findings reflect physicians' perceptions of error prevention, not actual reductions in medication error rates. The study emphasizes that while the model explains the perceived digital pathway, it does not measure objective clinical outcomes.

In conclusion, the study posits that for digital transformation to enhance patient safety in low-resource private clinics, the focus must shift from technology deployment to the socio-technical alignment of digital tools with the realities of clinical practice.