Exploring the feasibility of obtaining mifepristone and misoprostol from the internet

We searched Google using terms such as “buy abortion pills online” and “abortion pills online” to identify a convenience sample of websites that claimed to sell and ship mifepristone and misoprostol by mail to people in the United States without a prescription. We also obtained website recommendations from colleagues who had interviewed people researching their options for abortion medications online. Many sites offered multiple products advertised as containing both mifepristone and misoprostol, and these products were usually offered for approximately the same price. If a site offered multiple similar products, we opted to buy one called “MTP Kit” if available; products with this designation were the most commonly offered across sites. If an MTP Kit was not available, we purchased “Mifegest Kits.” From two sites, we purchased products advertised as containing only misoprostol.

Five researchers in four states (TX, WA, CA and NY) ordered pills and had them shipped to residential addresses in their state. Standard shipping was generally selected, except in four cases where express shipping was selected because the lower bound of the estimated standard shipping time was greater than 12 days. The researchers recorded information about the websites, price, payment, shipping, packaging and communications with vendors.

Researchers in states other than NY forwarded all packages received to NY by mail for storage at room temperature until we were ready to send them to a laboratory for chemical assay. At that time, we removed the tablets from each blister pack and placed the pills labeled mifepristone (if any) in one bag and all of the misoprostol pills in another, both labeled with their assigned ID letter. We did this separately for every product. The label did not reveal the identity of the website from which the pills were obtained. We mailed the bags to a testing laboratory that is registered and in good standing with the US Food and Drug Administration. The laboratory selected one tablet from each bag and assayed it for mifepristone or misoprostol chemical composition as labeled. The mifepristone assay used the reversed-phase high-performance liquid chromatography (HPLC) method used by this laboratory for testing of mifepristone tablets sold in the United States, adjusted for a single tablet. For the misoprostol assay, the laboratory developed a reversed-phase HPLC method and verified the linearity of analyte responses to confirm the suitability of the method for the single-tablet assay purpose. We assessed the relationship between the misoprostol content of the pills and the interval between package receipt and assay using a Pearson correlation coefficient.

The charge for the 18 mifepristone–misoprostol products received, inclusive of shipping and payment fees, ranged from $110.00 to $360.00 (median $204.50; Table 2). The two misoprostol-alone products were less expensive. The median shipping cost for all 20 products received was $25 with a range of $0.00 to $171.00; this included 4 express shipping charges and 18 standard shipping rates. Payment options included “E-Check” (a service that required the buyer to enter a bank account and routing number), wire transfer, PayPal, Visa, MasterCard, Western Union or Bitcoin, but sites rarely offered more than two or three of these options.

The interval between payment and package receipt ranged from 3 to 21 business days (median 9.5 business days; Table 2). Sixty-five percent of packages arrived within the advertised shipping time. Delivery of some packages may have been delayed because of winter weather at the buyer's location. One shipment was delayed by the vendor who believed that the order had been erroneous because the researcher had at the same time unknowingly also purchased a product from another website apparently run by the same vendor. According to package tracking information, 15 of the products received were shipped from locations within the United States, and 5 were shipped from locations in India. All sites provided a contact number or email address. We contacted most sites at least once between ordering and receiving the product to check on the status of the order or to confirm that the package had been shipped.

All pills were packed in blister packs made of foil on both sides. All 20 products received contained tablets labeled with the active ingredients ordered, but very few matched the photo used to advertise the product on the website. Most blister packages appeared intact (Table 3). Eight mifepristone–misoprostol products, all from the same manufacturer, had pinprick-sized holes in each of the five blisters, and the foil blister was indented. One product, which contained only misoprostol pills, arrived damaged: 11 of the 20 blisters split open, and some of the pills were broken. Only two of the blister packs came in boxes, both of which were printed with a brand name. None of the products came with instructions or any written communication. Each blister pack received had the brand name of the product, the amount of active ingredient per tablet (200 mg for the mifepristone pills and 200 mcg for the misoprostol pills) and expiration date stamped on the foil. None of the products were expired. All but one of the products had a suggested retail price (inclusive of taxes) stamped on the package, which ranged from 378.47 rupees to 690.00 rupees ($5.88 to $10.72 in US dollars). All of the products were labeled as having been manufactured in India. None of the shipping packages indicated that the contents were pharmaceutical products or related to abortion, and three came with other products in the envelope that were not ordered: two with cell phone charging cables and one with a necklace.

We found evidence that many of the sites were run by the same vendors. As noted above, one vendor admitted to having received two orders sent through different websites. Three purchases from different sites appeared on our credit card statement as having been paid to a single vendor. Wire transfers to pay for products from five unique sites were sent to bank accounts belonging to the same person. Order confirmations from three sites were sent from the same email address. Eight products we received were pinpricked in the same way, with the blisters folded around the pill. Finally, two products – ordered by the same researcher from different URLs days apart – were shipped to the researcher in the same shipping envelope.

During the ordering process, one website disappeared and then reappeared several days later. As of May 2017, 14 of the 18 sites ordered from were accessible online and 4 were not.

Two buyers received troubling communications from product vendors. One warned about the legality of purchasing online: “Please do not share this info with any other side because investigation team is searching the details for this type of medicine.” In the second case, the vendor complained that he was unable to get payment from the online payment platform and threatened to withhold shipment until the buyer paid another way. After some convincing, he finally sent the product. In addition, one buyer who paid with Western Union received two fraud alert calls — one from Western Union advising against purchasing pharmaceuticals online and another from her credit card company.

Assays for mifepristone were performed 8–84 days (median 40.5) after receipt of the pills from the vendor; the range for the misoprostol assays was 15–91 days (median 47). The analyses revealed that all mifepristone pills received contained mifepristone in amounts within 8% of 200 mg, ranging from 184.3 mg to 204.1 mg/pill (Table 3). All misoprostol pills assayed contained some misoprostol, but the amount per tablet varied from 34.1 mcg to 201.4 mcg. The eight pills with the lowest misoprostol content were the eight Pregnot brand products, manufactured by Akums Drugs & Pharmaceuticals Ltd., and were also the only eight with pinpricked packages. Misoprostol content was not correlated with the interval between package receipt and assay (p=.88). We ordered from four websites twice and found that pairs of pills from the same website had similar amounts of active ingredient.