It is usually a mistake to predict the future. Much evidence shows that predictions about technological advances are rarely correct. The experience with the economy in 2008 also shows that even overpaid bankers and academic economists are unable to predict the future even though it is their job and they have powerful mathematical tools to assist their predictions. However, you need certain assumptions about the future, which makes everyone an every day prophet. You might not purchase breakfast cereal and milk if you did not prophesy that there would be breakfast tomorrow morning.

A chilling recent article in Gastroenterology reports that mortality rates are still shockingly high in the United Kingdom for patients admitted to the hospital with gastrointestinal bleeding.

We used a case-control study design to analyze data from all adults administered to a National Health Service hospital, for upper gastrointestinal hemorrhage, from 1999 to 2007 (n = 516,153)…During the study period, the unadjusted, overall, 28-day mortality following nonvariceal hemorrhage was reduced from 14.7% to 13.1% (unadjusted odds ratio, 0.87; 95% confidence interval: 0.84–0.90). The mortality following variceal hemorrhage was reduced from 24.6% to 20.9% (unadjusted odds ratio, 0.8; 95% confidence interval: 0.69–0.95).

Based on these findings, innovative therapies to treat upper gastrointestinal bleeding and reduce bleeding from peptic ulcers and varices would be valuable. These therapies are likely to be in the form of vaccines or other methods to treat and manage infectious causes of gastrointestinal bleeding, especially chronic liver disease.

I rather like Arthur C. Clark’s first of three laws of prophesy, especially the first, which states, “When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is very probably wrong.” Clarke was a hugely successful science fiction writer who wrote the script for Kubrick’s film “2001: A Space Odyssey.” I will try to follow his advice to avoid saying that things are impossible.

Could gastrointestinal bleeding from peptic ulcer become a thing of the past? Of course not. (Oh dear, I have already broken Clarke’s first law of prophecy). Could a World Health Organization type of program eradicate Helicobacter pylori ? After all, it worked for smallpox. However, because Helicobacter is widespread and has animal hosts, multiple strains, and antibiotic resistance strains, and has required increasing use of quadruple antibacterial therapy, that this probably will not be possible. Progress in vaccine development has been slow.

One failure in Helicobacter studies has been the inability to show the epidemiology of transmission, which is usually the first step toward effective public health management. Fecal oral transmission seems likely. Flies on sewage have been considered. Could better strategies of cleanliness, filtration of the water supply, or different handling of sewage reduce transmission rates?

A recent review concluded that “decision analysis models suggest preventing acquisition of H pylori , via vaccination in childhood, could be cost-effective and may reduce incidence of gastric cancer by over 40%. As yet, no country has adopted public health measures to treat infected individuals or prevent infection in populations at risk.”

Removing the handle from a cholera-infected well in London in 1854 (see Appendix 1 ), based on the epidemiologic mapping studies of John Snow, led to the end of a lethal outbreak and then control of cholera in most of the world. If the corresponding handle-limiting Helicobacter transmission could be identified and removed, this would be beneficial.

Could immunization against Helicobacter markedly reduce its incidence, with a consequential reduction in the incidence of bleeding peptic ulcers and gastric cancer?

A recent review of progress in Helicobacter vaccine development concluded that: …several key bacterial factors have been identified: urease, vacuolating cytotoxin, cytotoxin-associated antigen, the pathogenicity island, neutrophil-activating protein, and among others. These proteins, in their native or recombinant forms, have been shown to confer protection against infectious challenge with H pylori in experimental animal models. It is not known, however, through which effector mechanisms this protection is achieved. Nevertheless, a number of clinical trials in healthy volunteers have been conducted using urease given orally as a soluble protein or expressed in bacterial vectors with limited results. Recently, a mixture of H pylori antigens was reported to be highly immunogenic in H pylori-negative volunteers following intramuscular administration of the vaccine with aluminium hydroxide as an adjuvant.

The incidence of Helicobacter and its related diseases—duodenal and gastric ulcer, gastric lymphoma, and especially gastric cancer—seems to be decreasing, although this decline is probably unrelated to intentional medical intervention.

Could immunization against hepatitis C reduce variceal bleeding? Vaccination against hepatitis B has successfully reduced hospital admissions from bleeding caused by this virus and probably also reduced variceal bleeding. The hepatitis C virus infects at least 170 million people worldwide and approximately 4 million people in the United States. It is a significant public health problem because most acute hepatitis C infections become chronic, which can lead to further liver problems, such as cirrhosis and cancer. A hepatitis C vaccine would be a great victory for preventative medicine. The technical problems are formidable because the hepatitis C virus has developed ways of evading the host’s immune response to establish persistent infection. Vaccines in clinical trials now include recombinant proteins, synthetic peptides, virome-based vaccines, tarmogens, modified vaccinia, Ankara-based vaccines, and DNA-based vaccines.

Patients with AIDS and HIV can have upper gastrointestinal bleeding from a wide variety of causes. In 2010, approximately 35 million people out of a world population of 7 billion are now infected with HIV. Could immunization against HIV reduce acute admissions from gastrointestinal bleeding in these patients? Funding will probably be provided for further developments, and the heartbreakingly modest progress in this area will likely continue. Trials with vaccines have started. A clinical trial involving 16,402 subjects in Thailand, which represented a joint project between the governments of Thailand and the United States, provided the first demonstration that an AIDS vaccine can protect humans from HIV infection. Although the vaccine candidate tested was only modestly effective, it perhaps provides researchers a platform on which to improve.

Multiple genomic virus types and rapid viral adaptations, in contrast with slow vaccine developments, may allow viruses to stay ahead of immunizations methods. Technical advances in rapid vaccine developments have recently been shown to be possible in response to H1N1 swine flu; innovation in this area could be important.

Other common infectious causes of upper gastrointestinal bleeding may be amenable to innovations in vaccine development. Dengue is a common tropical disease caused by a mosquito- or tick-borne arbovirus that is increasing in incidence (50–100 million infected yearly in 110 countries). A proportion of people affected, commonly children or young adults, experience serious gastrointestinal bleeding secondary to dengue hemorrhagic fever, and die. Current treatments are poor, consisting of fluid replacement and transfusion, and no vaccines have been approved. Effective immunization would be valuable. Phase 1 human trials of a vaccine were recently reported.

Drug-induced upper gastrointestinal bleeding seems almost certain to increase in frequency as medication are increasingly used in the hope of maintaining health. Research in arthritis has drifted away from NSAID development towards more lucrative monocolonal antibody drugs of murine, chimeric, humanized and human types. NSAIDS are likely to continue to be used in large volumes because they are cheaper than monoclonals and may not change in terms of their risk of causing gastrointestinal bleeding. Aspirin seems to be irreplaceable for protection against stroke and heart disease, and may improve the outcome of colon, lung, and breast adenocarcinoma, perhaps through inhibiting PTGS2 (COX 2) enzymes. If anything, aspirin will probably be used with increasing frequency, especially in the third world, because it is inexpensive. Use of over-the-counter nonsteroidal anti-inflammatory drugs will probably increase the incidence of gastrointestinal bleeding worldwide.

Clinicians treating cardiovascular disease, stroke, and clot formation often use medications that reduce coagulation in different ways. Aspirin induces erosions and reduces platelet function. Drugs such as warfarin and the increasingly used drug clopidogrel also increase bleeding but probably do not alter ulceration rates. An increase in drug-related gastrointestinal bleeding rates will probably occur. If more patients with arthritic, cardiac, and cerebrovascular diseases are helped by the new (or old) drugs than are harmed by occasional gastrointestinal bleeding complications, this may need to be a grudgingly accepted trade-off.

Could the incidence of alcohol-related variceal hemorrhage be substantially reduced through health initiatives? An analogy might be the observation that campaigns against smoking have altered smoking habits in many Western countries. The absence of smoking that has been achieved in the underground in London and in restaurants and pubs would have been unthinkable a few years ago and has probably saved lives. Smoking remains a factor in the origin of peptic ulcer bleeding. Tobacco causes 650,000 deaths in the European Union each year. It is still the single largest cause of death, disease, and disability. The surprisingly successful recent marginalization of smokers that has occurred in several Western countries, with extension of smoke-free habitats if applied to the rest of the world, will probably reduce the incidence of bleeding ulcer. Reduction of alcohol intake might be achievable, but banning alcohol in restaurants and pubs and bars seems unlikely, and almost unthinkable, unfortunately. The prohibition of alcohol in the United States during the 1930s is not regarded as a great success.

Could better drugs be available for the treatment of gastrointestinal bleeding? Some people think that the use of proton-pump inhibitors has improved outcomes after admission for upper gastrointestinal bleeding. Some randomized evidence supports this, although the effect is probably fairly small; large well-conducted trials of H ₂ antagonists did not show improvements in outcome after admission. Trials in bleeding peptic ulcer are difficult to conduct because large numbers of patients are needed. Studying multicenter groups would be an improvement, but this requires an organizer with genius and tact, and access to a lot of money. If competition from a drug or drugs that had a possible edge on omeprazole existed and the manufacturers wanted to establish an advantage, this would help. This scenario might lead to funding of well-conducted large-scale trials with adequate blinding and monitoring.

Progress might be seen with combination drug therapy for ulcer bleeding. A proton-pump and a protease inhibitor might be worth considering. The problem with bleeding ulcers is that a hole is present in the wall of an artery.

Could drugs inhibiting angiogenesis have any role in the management of bleeding from telangiectasias and angiomas? These lesions are the commonest cause of bleeding from the small intestine and are currently poorly managed with endoscopic therapy and surgery.

Somehow drug therapy seems to be a poor first-line medical response to acute bleeding from ulcers or varices. Bleeding from arteries at other sites in the body than the gut would not be treated with drug therapy if a more sensible way existed to stop the bleeding mechanically.

What innovations seem likely in endoscopic diagnosis? Currently, patients presenting in the emergency room with gastrointestinal bleeding are still assessed with a nasogastric tube, which is a rather barbaric approach that is used widely without much diagnostic value. Clinical scoring systems (Forrest, Baylor, Rockall) may be of some value in predicting rebleeding or mortality but could be improved if better anatomic information about the state of the bleeding artery or vein was available or if rebleeding could be predicted more accurately. Doppler probes have had occasional advocates. Endoscopic ultrasound might be able to give better information about the anatomy of the eroded vessel in an ulcer or the distribution and pressure in bleeding varices, although this technique would need to become a part of out-of hours and emergency endoscopic practice.

A wireless capsule that uses real-time imaging that cane be administered by nursing staff in the emergency room might be worth exploring. Although this technique would be more expensive than a nasogastric tube, it would probably have much more diagnostic value. If it reduced the numbers of patients requiring emergency endoscopy and inpatient treatment, then it might have cost advantages and may allow triaging of patients most in need of endoscopic therapy, emergency surgery, and intensive care–style observation. Recent advances with remote magnetic manipulation in the stomach might offer improved visualization. The fact that wireless capsule endoscopes could be administered by staff without flexible endoscopic skills also might have cost advantages and reduce out of hours endoscopy by skilled staff.

Another way to use capsules would be to attach one to the wall of the stomach to observe whether further bleeding occurs. The frame rate might need to be altered, because rebleeding tends to be common in the first 2 days and to decrease in frequency exponentially thereafter. It would be helpful if the capsule could be switched on during ward rounds or by nursing staff if a patient’s pulse-rate increases or the blood pressure drops. The ability to see the disappearance of stigmata of bleeding with time would also be reassuring.

Could capsules be made to stop bleeding? A capsule that includes an endosurgical generator would be difficult to make; it is thinkable (which is my code for nearly impossible, and therefore an attractive but high-risk area for innovation). The power available in two 3V batteries could deliver a few pulses of energy sufficient to burn tissue, but current battery structure and internal resistance make it difficult to get enough power out of these batteries to burn tissue despite advances in voltage amplifiers and capacitor storage size diminutions.

Could telemetry assist in the diagnosis and management of upper gastrointestinal bleeding? Although medical telemetry has its advocates, the pace of nonmedical technological advances in telemetry continues to make a mockery of the special requirements for high-cost medical telemedicine technology. Special expensive high-definition video cameras and monitors are not needed for most medical imaging purposes, because they have become cheaply available for nonmedical applications. Neither are expensive dedicated medical teleconferencing facilities needed, because Skype is free. However, a sensible, economic, nonrestrictive importation of technology into front-line medicine seems oddly difficult to achieve.

Earlier telemetric signaling of rebleeding, whether from an attached capsule or from machines measuring blood pressure and pulse, that bypass the nursing station, might be useful. The usefulness will be contingent on having more-effective and less-invasive treatments for patients with continued or recurrent bleeding, and on sensible decisions made by on-call staff, with more information reaching them through phones or computers.

Real advances have been made in the management of arterial bleeding in laparoscopic and open surgery. The harmonic scalpel and bipolar forceps, with the steady improvement in stapler design, have enabled bleeding from intra-abdominal or thoracic arteries and veins to be much less troublesome.

Could these advances be applied through flexible endoscopes? Bipolar forceps, but currently not harmonic scalpels, are easy to make and use during flexible endoscopy but have not yet attracted much attention as an important innovation in endoscopic hemostasis. Some Japanese endoscopists performing endoscopic mucosal resection and endoscopic submucosal dissection have used bipolar forceps for hemostasis. Although flexible bipolar forceps can undoubtedly stop bleeding from large vessels in the gastrointestinal tract, whether this technology can stop bleeding from large bleeding vessels in the floor of ulcers is unknown. Sprays or topical applications of recombinant thrombin, fibrinogen, and thrombin mixtures, cyanoacrylates, and nano-SiO ₂ are currently fashionable for some surgical hemostasis, including reducing bleeding needle puncture sites in bleeding aortic aneurysms. Some of these ingredients have been used in endoscopic hemostasis, but their efficacy is difficult to assess relative to thermal and mechanical methods, such as bipolar probes or clips. Ease of application is no substitute for hemostatic efficacy. The vessel in the base of a bleeding ulcer may be aneurysmally dilated, and covering it with a hemostatic film might be a good idea. Most surgeons would prefer to tie the vessel above and below the bleeding point and to place the thread around tissue that is a little distance from the damaged base of the ulcer. Getting underneath the bleeding vessel would involve a full-thickness perforation in approximately half of bleeding ulcers. That procedure would have been unthinkable a few years ago, but closing iatrogenic perforations has become almost commonplace after the development of endoscopic submucosal dissection. The force required for a surgeon to pass a curved Mayo needle into a sclerotic duodenal ulcer and around an eroded gastroduodenal artery to tie it in a recommended figure-of-eight configuration is considerable. That action would be difficult to mimic effectively using instruments passing through a flexible endoscope.

Doctors tend to overestimate the efficacy of the treatments they prescribe and to be pompous about the quality of the evidence that supports their favorite therapies. An injection of realism into endoscopic therapy for bleeding ulcers would be a valuable innovation. Recognizing that current endoscopic therapy methods of hemostasis are poor is an important first step toward improving them.

Injection of adrenalin therapy as primary therapy for bleeding ulcers (and injection of sclerosants for bleeding varices) seems to be waning. These therapies were popular because they were inexpensive and widely available. Injection therapy is ineffective at stopping bleeding in animal models. Surgeons would not use injection if they could tie, staple, or coagulate a bleeding ulcer with efficient coaptation (squeezing). Small powered trials were unable to show that injection was less effective than thermal probe methods. Bipolar probes are better than injection methods in models of bleeding ulcers in animals but in humans are probably only of modest efficacy in vessels more than a millimeter in diameter at the base of a bleeding or visible ulcer. It is hard to occlude a piece of tubing on a bench, never mind an artery that is running in the bed of an ulcer with a 2.8- or 3.2-mm bipolar probe passed through a flexible gastroscope. The ability to place forceps on both sides of a bleeding artery and squeeze it is why surgical hemostasis with bipolar forceps or a harmonic scalpel is so effective. Most endoscopic clips in their current designs are of limited efficacy because they have a gap and therefore do not effectively compress small arteries. They are also conceptually limited in that they are designed to either pass through small channels of gastroscopes or be mounted on the outside of the endoscope and released using a band ligation–like delivery.

Clips of a design almost identical to those used today were first used through flexible endoscopes in Japan in 1971 and were probably the first device to be used at flexible endoscopy to treat bleeding. Injection sclerotherapy was used by a few pioneers in rigid endoscopy to treat bleeding varices before that time. Clips that are bigger and of different designs that make them easier to load and use seem to be an obvious area for development and carry my vote as an innovation with the greatest potential impact in flexible endoscopic hemostasis.

One advancement that has probably been helpful has been educating endoscopists on how to use clips. The clips used to be hard to load and impossible to rotate, and clinicians found it difficult to remember the right sequence to fire them successfully. When substantial training is required to perform an endoscopic procedure, this suggests that the procedure or device needs improving. Some improvements have been made in the delivery of clips.

Regarding ulcer excision and closure, the experience with natural orifice translumenal endoscopic surgery (NOTES) showed that both are possible using flexible endoscopes. The tools that make these procedures fairly easy to perform have not become commercially available. Many of the tools that made NOTES possible would be valuable during intralumenal flexible endoscopy but will not be available because the flexible endoscopic market size is small. The instruments were mostly developed by laparoscopic surgical companies, which do not have good access to flexible endoscopic market outlets.

Can endoscopic innovation improve the management of variceal bleeding? The fact that drug therapy, and recently even radiologic shunting, seems to be overtaking band ligation as a method for managing varices suggests that there is room for an innovative flexible endoscopic mechanical surgical method that is superior to band ligation. Transoral stapling seems to be one possible answer to effective full-thickness internal devascularization of esophageal varices. It seems strange that flexible endoscopic staplers have taken so long to develop and have not become commercially available. Sengstaken and Linton balloons are still used for tamponade in some patients with unresponsive variceal bleeding. Their use seems barbaric, and alternative methods of tamponade seem ripe for technical innovation.

Increasing red tape and inefficient regulation are likely to prolong the time and expense required for introduction of any innovative treatment.

The adverse impact of economic recession on medical innovation is an important factor reducing the rate of change. It seems arguable, although perhaps naïve, to think that the forces driving medical innovation are more from dissatisfaction with current treatments than from the desire for financial gain through the exploitation of a new idea. Without venture capital, very few innovations stand much of a chance of ever entering mainstream medical therapy.

The recent economic recession in the United States has caused a fourfold reduction in the rate of venture capital investment, from a peak of $40 billion in 2007 to $11 billion in 2010. Higher-risk ventures, which will include some really useful innovative medical developments, are much less likely to be funded than those that seems safely similar to less-innovative concepts. A marked reduction is likely to be seen in funding of the development of innovative flexible endoscopic treatments and development of drugs, some of which might help reduce gastrointestinal bleeding. The continuation of the economic downturn will also reduce the rate of development of cardiovascular and arthritis drugs that have unintended gastrointestinal complications; this might slow the steady increase in iatrogenic causes of gastrointestinal bleeding.