Prediction Markets (PM), also known as Information Markets (IM) or Event Futures, are mechanisms used to predict future outcomes by harnessing the collective knowledge of participants. This is achieved through trading contracts related to specific events that require predictions. Participants in these markets receive payoffs based on the contracts they hold and the accuracy of the future outcomes they predict. ⁵ The University of Iowa pioneered the use of Prediction Markets (PM) with the establishment of the Iowa Electronic Market (IEM) in 1988. This PM was designed to forecast the outcome of the US presidential election and involved the creation of trading contracts. If a participant’s election prediction proves accurate, they receive the agreed-upon payout through this electronic prediction market. ^{6, 5}

In fields such as pharmaceuticals and technology, PM can be used to forecast the success of research and development projects. This helps allocate resources more effectively. Graefe’s study applied the PM to predict whether a new drug to be developed will pass future clinical studies. The study found that the effectiveness of the PM depends on 1) the participant is relevant and knowledgeable in the relevant subject; 2) the participant is willing to provide useful information to the research study; and 3) the investigator. Able to manage the information received properly and appropriately. ⁷

PM is used in various ways; therefore, it is important to understand the terminology associated with PM, as in Table 1.

Previous studies have demonstrated the predictive accuracy of PM. When compared to the Delphi Method, it was observed that the Delphi Method was user-friendly and cost-effective. PM’s versatility makes it applicable across various industries and forecasting scenarios due to its user-friendly nature. Additionally, PM can operate continuously and incentivize participants to provide accurate predictions based on their genuine beliefs. Nonetheless, participants are required to possess knowledge of marketing and contract trading for effective engagement. ⁴

When it comes to forecasting, the PM method is as precise as opinion polls. In predicting the percentage of support for US presidential candidates, it exhibited a mere 1.5% margin of error, whereas opinion polls had an error rate as high as 2.1%. ⁶

When comparing face-to-face meetings, nominal groups, the Delphi methodology, and Prediction Markets (PM) across ten different questions, it was discovered that the Delphi Method outperformed direct face-to-face meetings, while the PM and small group meetings yielded similar levels of accuracy. However, participants expressed lower satisfaction with the PM due to its perceived limitations, including difficulty in access and lack of interactivity, attributed to its computerized nature. ⁸

Kloker et al proposed the integration of a prediction market in conjunction with Delphi studies and the selection of participants based on their trading behavior within the market for the Delphi study. This study introduces attributes of trading behavior that are theoretically derived from existing literature and may be associated with informed traders. These attributes are subjected to testing and evaluation using historical prediction market data. In terms of originality and value, the identified attributes can be harnessed to make the selection of experts for Delphi studies more objective, potentially leading to increased consideration of information in the process. ⁹

This study employed the PM process for forecasting, as PM has proven accuracy in predicting future events related to the development of dosage forms for IMDs. This study created a Google Sheets-based application to facilitate participation in PM and simplify the prediction process. Furthermore, while traditional marketing forecasting primarily focuses on future predictions, our research sought to provide a rationale behind these predictions. Hence, we integrated the PM process, allowing participants to provide justifications for their choice of trading contracts concerning specific dosage forms in each round of the contract initiated.

In Thailand, Sripawatakul proposed a four-step decision flow diagram for creating a PM, outlined as follows ^{10– 12} : 1.

Select a matching mechanism for the trading contract from the four available options: Continuous Double Auction (CDA), Pari-Mutuel, Dynamic Pari-Mutuel, and Market Scoring Rule (MSR).

The objective of this study was to identify new dosage form developments for IMDs for the domestic pharmaceutical industry in Thailand using Prediction Markets (PM). The literature review on new dosage form development covered eight types: sustained release, orodispersible tablets, soft gel capsules, delayed release, transdermal patches, nasal sprays, thin films, and powder sachets. 1.

The study chose the Winner-Take-All type of prediction market (PM) because it is easy for participants to understand and aligns with the preferences of many market makers. In a Winner-Take-All contract, each contract is priced at p baht, and if a participant accurately predicts the event’s outcome, they receive a payoff of 1 baht. The contract price represents the associated probability.

“ B” serves as a constant determining the upper limit on the payout amount market makers can offer.

“ q 1 ” represents the quantity of contracts purchased for outcome 1.

“ q n ” represents the quantity of contracts purchased for outcome n. 3.

The prediction market (PM) application was conducted using Google Sheets and featured eight dosage form options: sustained release, delayed release, orodispersible tablet, nasal spray, chewable tablet, thin film, transdermal patch, and powder in a sachet. These forms were identified through literature reviews and input from pharmaceutical industry experts and prescribers. Participants could specify the number of contracts or shares they wished to purchase for a particular dosage form by selecting the corresponding cell. The initial share or contract price was determined using a starting pricing formula, which was then adjusted based on market demand in subsequent rounds. The study consisted of four rounds, and investors were encouraged to keep investing until information saturation was reached.

The research was approved by the Human Research Ethics Committee of Khon Kaen University (HE642276) on January 20, 2022. Each participant was required to complete a consent form by mail after receiving a thorough explanation of the study’s objectives and methods for trading contracts prior to participating in the prediction market. Participants were given a confidential investment code that was not shared among them.

We specifically invited participants who are executives or decision-makers with the authority to make investment decisions in the domestic pharmaceutical manufacturing industry and are members of the Thai Pharmaceutical Manufacturers Association (TPMA). Those who provided consent participated in an online Zoom session lasting approximately 60 minutes. During this session, audio was recorded for subsequent data analysis. It was emphasized that these recordings would be securely deleted upon the study’s completion.

The prediction market experiment was performed using google sheets. Rules and method to trade the contracts were instructed as following: 1.

The primary question in the prediction market experiment was, “Which type of dosage form will the domestic pharmaceutical industry successfully invest in for research and development?” Participants expressed their opinions by buying stocks in various dosage form candidates: sustained release, delayed release, orodispersible tablet, nasal spray, chewable tablet, thin film, transdermal patch, and powder in a sachet. The number of stocks purchased indicated the perceived likelihood that these dosage forms should be prioritized for research and development.

A total of 28 participants accepted the invitation to join the PM experiment. Regarding their job positions, the participants can be classified into five categories: Executive Director, R&D, Marketing, Production, and Registration. The majority of participants were executive directors as shown in Figure 1.

Total dosage forms selected by participants across 3 rounds are detailed as follows:

The drug development directions were obtained from the drug registration and approval information available on Drugs@FDA: FDA-Approved Drugs ( www.accessdata.fda.gov). ¹⁶ We identified 10 groups of high-value drugs commonly used in Thailand. The most developed dosage forms across all treatment groups were the sustained release, orodispersible tablet, and delayed-release tablet respectively. In this analysis of the prediction market, the identical outcomes from the searching were sustained-release dosage forms, followed by orodispersible tablets.

The study’s findings revealed a total of 8 dosage forms, which include sustained release, delayed release, orodispersible tablet, nasal spray, chewable tablet, thin film, transdermal patch, and powder of sachet. The domestic pharmaceutical industry or participants for PM, predicting “What is your prediction for the type of dosage form that the domestic pharmaceutical industry will successfully invest in and develop?” As a result, 3 dosage forms are anticipated for future development: sustained release, orodispersible tablet, and nasal spray.

An examination of the accuracy of PM results for predictive purposes in the field of dosage form development of IMDs was conducted as follows: Comparative analysis of PM results and evaluation assessment reports from 505(b)(2) ¹⁷ USFDA particularly in the issue of dosage form development for IMDs. It was found that for most of the drugs after USFDA approval and entering the market, the dosage form of the IMDs was sustained release. ¹⁶ Similar to the findings of in-depth interviews with prescribers in the previous study, it was found that they prefer the development of SR tablets to improve medication compliance and reduce the frequency of medication intake. ¹⁸

In 2020s, the growth rate for locally produced new generic drugs was relatively modest, with a Compound Annual Growth Rate (CAGR) of 7.12%. In contrast, the situation was distinct for imported new generic drugs, displaying a higher growth rate at a CAGR of 16.21%. ¹⁹ Nonetheless, the latest generic medications still encompass over 54% of the market share. Therefore, this study is advisable for Thailand to prioritize the advancement of this sector to encourage the development of SR tablets for IMDs. In this research, the majority of PM participants possess the capability to develop new generic drugs as well as SR and other dosage forms.

In the realm of PM applications, numerous options have emerged since the inception of the Iowa Electronic Market ²⁰ However, participant satisfaction with these platforms has often been marred by difficulties in access and a perceived lack of interactivity due to their computerized nature. ⁸

Consequently, the research team developed a user-friendly Excel-based PM application, drawing inspiration from the concept of the Iowa Electronic Market. When utilizing PM, it is advisable to consider the following factors: 1.

The future event to be predicted should possess a clearly definable outcome to facilitate accuracy measurement.

The incentive behind this Prediction Market (PM) was the prospect of participants receiving the study results if they secured ranks within the top three positions. Previous research has identified two primary categories of motivations: 1) Ranking-based incentives and 2) Performance-driven incentives. Notably, a study conducted by Kridchapat in 2012 demonstrated that ranking-based incentives consistently outperformed performance-driven incentives in terms of prediction accuracy. ¹⁴

Therefore, this paper presents two key contributions. Firstly, it stands as the inaugural investigation into the PM study of dosage form advancement for IMDs in Thailand, culminating in the forecast of sustained-release dosage forms. Should the domestic pharmaceutical industry embrace this prognosis and embark on the development of sustained-release dosage forms utilizing new or advanced technologies, it could significantly augment its competitive edge on the global stage. Secondly, the paper proposes a roster of developed dosage forms across ten treatment categories: lipid-modifying agents, medications for peptic ulcers and gastroesophageal reflux disease (GERD), primarily vascular-effect-focused selective calcium channel blockers, blood glucose-lowering drugs, NSAIDs, beta-lactam antibacterial agents, antiepileptics, systemically used antihistamines, antipsychotics, and beta-blocking agents.

The PM through the order-matching mechanism uses the price function method to predict the patterns that the domestic pharmaceutical industry will be able to produce IMDs which has executives or decision-makers to invest in the domestic pharmaceutical industry. The first three predicted outcomes were sustained release, orodispersible tablet, and nasal spray.